If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Women´s and Children´s Health, Karolinska Institute, and the WHO Collaborating Centre for Research in Human Reproduction, Karolinska University Hospital, Stockholm, Sweden
Department of Women´s and Children´s Health, Karolinska Institute, and the WHO Collaborating Centre for Research in Human Reproduction, Karolinska University Hospital, Stockholm, Sweden
Levonorgestrel-only emergency contraception (EC) inhibits ovulation to prevent fertilization. Misconceptions regarding its mechanism of action contribute to low use in some settings. We aimed to review the mechanism of action of Levonorgestrel EC and assess the evidence for a postovulatory effect on viable pregnancy development in the exposed and subsequent cycles.
Basic procedures
We searched Pubmed, Embase, Web of Science, clinicaltrials.gov, the Cochrane database, and the US FDA Adverse Event Reporting System. We included studies reporting on the effect of Levonorgestrel EC exposure on (1) ovulation, (2) conception, (3) implantation, (4) ectopic pregnancy, (5) pregnancy viability, and (6) fertility in subsequent cycles. Data were extracted and tabulated from included studies by two researchers. The results were analyzed and summarized in narrative form.
Main findings
We included 33 studies. Women exposed to Levonorgestrel EC at or after the LH (luteinizing hormone) surge had similar rates of inhibited ovulation (3 studies) and conception rates (2 studies) as controls. Nine out of ten studies found no difference in endometrial receptivity to implantation after exposure to Levonorgestrel EC compared to controls. The evidence for an association between Levonorgestrel EC and reduced fallopian tube motility or ectopic pregnancy was conflicting. We found no evidence of increased miscarriage or teratogenicity (3 studies) or disrupted menstrual cycle (3 studies) after Levonorgestrel EC exposure compared to controls.
Principal conclusions
This review supports that Levonorgestrel EC taken after ovulation does not affect implantation and results in similar conception rates compared to placebo. There is no evidence that exposure to Levonorgestrel EC affects fetal development, miscarriage, stillbirth, or subsequent menstruations.
]. It was introduced in the 1990s, and Food and Drug Administration approved based on an estimated efficiency of 84% when taken within 3 days of unprotected sex [
]. Systematic literature reviews and reports of controlled trials have shown that adverse events related to Levonorgestrel EC use are rare, reported side effects appear to be mild, and acceptability among women is high [
A prospective, open-label, single arm, multicentre study to evaluate efficacy, safety and acceptability of pericoital oral contraception using levonorgestrel 1.5 mg.
PATH, the World Health Organization, and the United Nations Population Fund. Essential Medicines for Reproductive Health: Guiding Principles for Their Inclusion on National Medicines Lists. Seattle: PATH; 2006.
Key statistics from the National Survey of Family Growth . Centers for disease control and prevention. Atlanta: National Center for Health Statistics; 2017.
Use of non-emergency contraceptive pills and concoctions as emergency contraception among Nigerian University students: results of a qualitative study.
]. The narrative that EC affects future fertility is exemplified by a survey in which 4 in 10 women believed that repeated use of EC could make them infertile and half of them thought that it could only be used once per cycle.[
Our aim was to review the mechanism of action of Levonorgestrel EC and assess the existing evidence for (1) the postovulatory effect of Levonorgestrel EC exposure on viable pregnancy development in the exposed cycle, and (2) the effect of Levonorgestrel EC exposure on fertility in subsequent cycles, compared to controls.
2. Methods
2.1 Literature search
We performed searches in Pubmed, Embase, Web of Science, clinicaltrials.gov, the Cochrane database, and the US Food and Drug Administration Adverse Event Reporting System (FAERS) Public Dashboard up until January 15, 2021. No other limitations were made regarding language, setting, or study period. A search module relevant to each study question was designed for each database. The search model used can be seen in Supplementary Table 1.
2.2 Population
The study population was women using Levonorgestrel EC or cell line or cultured human tissue exposed to levonorgestrel in vitro. Our comparison group included women, human cell-lines or tissue unexposed to levonorgestrel EC.
2.3 Intervention
We defined the intervention as exposure to Levonorgestrel EC at doses equivalent to those used for EC, i.e., single oral or vaginal administration of 1.5 mg, or two split doses of 0.75 mg at a 12-hour interval, or either of these doses taken recurrently at any time during the menstrual cycle. We defined the control population as women not exposed to Levonorgestrel EC, or the same woman unexposed to Levonorgestrel EC in another cycle, where women acted as their own control. In the case of in vitro studies, cell line or human tissue was treated with Levonorgestrel EC at pharmacological or supra-pharmacological concentrations, with the control population defined as biological tissue not exposed to levonorgestrel.
2.4 Outcomes
Our outcomes of interest in relation to Levonorgestrel EC effect in the exposed cycle were: (1) ovulation, (2) conception, (3) implantation, (4) ectopic pregnancy, and (5) pregnancy viability.
We defined inhibited ovulation as delayed or inhibited LH surge ≥5 days, or absent ovulation verified by ultrasound. As a measure of conception rate, we accepted the verification of a pregnancy by a pregnancy test or ultrasound. As measures of reduced implantation, we included measures of embryo-endometrial attachment rate or changes in biomarkers in endometrial constructs cultured in vitro. As measures of ectopic pregnancy, we accepted a clinical diagnosis of ectopic pregnancy by serial serum-hCG measurements, ultrasound, and/or surgery. We also accepted histological changes or biomarkers consistent with reduced fallopian tube motility. Tubal motility is relevant to the transport of the fertilized ovum to the uterus for implantation and reduced motility could theoretically increase the risk of ectopic pregnancy and non-viable pregnancy. We did not include data on outcomes that preceded fertilization such as sperm motility or cervical mucus changes in the exposed cycle as these precede fertilization.
As measures of pregnancy viability, we accepted the following outcomes: increased rates of miscarriage, stillbirth, or fetal abnormalities in pregnancies conceived in the exposed cycle.
Our outcomes of interest in relation to the impact of Levonorgestrel EC on future fertility were the following: (1) resumption of a normal menstrual cycle or ovulation in the subsequent cycle, (2) increased use of assisted reproduction treatment or reported infertility post-exposure, (3) time to conception in subsequent cycles, and (4) fertility within 12 months of exposure.
2.5 Data extraction
We included original research articles on the association between the use of Levonorgestrel EC and one or more of our outcomes of interest. Identified citations were uploaded into an EndNote database (EndNote X9) and duplicates were removed. Two researchers (ME and RHWL) screened citations by title and performed a full-text or abstract analysis of reports seeming to fulfill inclusion criteria. We excluded studies on other forms of EC. We excluded case reports, opinion pieces, or review articles.
2.6 Data analysis
To differentiate between pre- and post-ovulatory effects we presented outcome measures according to the timing of Levonorgestrel EC exposure: before the LH surge corresponding to a time that allows inhibition of ovulation, or at or after the LH surge corresponding to a time when ovulation cannot be inhibited. Outcomes were standardized for comparability if possible. We standardized the timing of levonorgestrel administration in relation to ovulation to refer to days (+/-) from the LH surge or follicle size. In vitro levonorgestrel concentrations were converted from μg/L to μM and related to serum concentrations after oral Levonorgestrel EC intake. Peak serum concentration of levonorgestrel after a single 1.5 mg oral dose is 12.3 ng/mL or 0.039 μM [
]. Data were summarized as average incidence and/or incidence range. The data were tabulated and presented in descriptive form. The research did not require ethical approval.
2.7 Quality of evidence
We did not apply a systematic grading of the quality of evidence, but we assessed strengths and weaknesses of included studies based on (1) the existence of a placebo (or equivalent controlled) group, (2) power (sample size), and (3) accuracy of the timing of exposure in relation to ovulation. Placebo (or equivalent controlled) studies with adequate sample size and reliable timing of exposure in relation to ovulation, verified by ultrasound and/or serum-hormone measurement, were assessed as high-quality evidence. Placebo (or equivalent controlled), with small sample sizes but reliable timing of exposure in relation to ovulation, were assessed as moderate-quality evidence. Uncontrolled studies with unreliable or poorly described timing of exposure in relation to ovulation were assessed as low-quality evidence.
3. Results
Our search identified 1325 articles that were screened by title or abstract. We included 33 studies in the review that reported on one or more of the following outcomes: ovulation (n = 9), conception rate (n = 2), fallopian tube motility (n = 5), implantation (n = 10), ectopic pregnancy (n = 3), pregnancy viability post-exposure (teratogenicity, miscarriage or stillbirth, n = 3), and return to normal menstrual cycle (n = 3). We identified no studies on increased use of assisted reproduction treatment or reported infertility post-exposure, time to conception in subsequent cycles, or fertility within 12 months of Levonorgestrel EC exposure.
A flowchart illustrating the inclusion process is shown in Figure 1.
Fig. 1Prisma Flowchart of inclusion process in a systematic review of use levonorgestrel for emergency contraception (LNG-EC) and fertility after failed use and in subsequent cycles.
Ulipristal acetate prevents ovulation more effectively than levonorgestrel: analysis of pooled data from three randomized trials of emergency contraception regimens.
Effect of a single vaginal administration of levonorgestrel in Carraguard (R) gel on the ovulatory process: a potential candidate for "dual protection" emergency contraception.
Pituitary-ovarian function following the standard levonorgestrel emergency contraceptive dose or a single 0.75-mg dose given on the days preceding ovulation.
Among studies reporting on the efficacy of Levonorgestrel EC in relation to the timing of ovulation, 5 were placebo-controlled (or equivalent) and of moderate to high-quality evidence [
Ulipristal acetate prevents ovulation more effectively than levonorgestrel: analysis of pooled data from three randomized trials of emergency contraception regimens.
]. Three of these studies found that Levonorgestrel EC had a significant inhibitory effect on ovulation compared to placebo or controls when taken at a follicle diameter of 15 to 17 mm or 2 days prior to the LH surge. The rate of inhibited or delayed ovulation ≥5 days was 71% vs 0% (p = 0.03), 91% vs 45% (p = 0.003), and 83% vs 17%, respectively (p < 0.05) [
Effect of a single vaginal administration of levonorgestrel in Carraguard (R) gel on the ovulatory process: a potential candidate for "dual protection" emergency contraception.
Pituitary-ovarian function following the standard levonorgestrel emergency contraceptive dose or a single 0.75-mg dose given on the days preceding ovulation.
]. No study found any significant effect on ovulation for Levonorgestrel EC compared to placebo when taken at follicle size ≥18mm. Inhibited ovulation varied between 15% and 50% for Levonorgestrel EC compared to 4% to 13% for placebo [
Ulipristal acetate prevents ovulation more effectively than levonorgestrel: analysis of pooled data from three randomized trials of emergency contraception regimens.
]. One study included a sub-analysis of the rate of inhibited ovulation at the start of the LH surge (n = 18), before the peak (n = 24), and at the peak (n = 46) and saw no significant difference compared to placebo in any of these groups [
Ulipristal acetate prevents ovulation more effectively than levonorgestrel: analysis of pooled data from three randomized trials of emergency contraception regimens.
Table 1Summary of original research studies reporting on ovulation rates or conception rates after exposure to Levonorgestrel for emergency contraception
Study year
Study design
Sample size
Treatment timepoint and outcomes
Treatment outcome: inhibited, delayed or afunctional ovulation
Grading of evidence: A= Assessed as high-quality evidence. Placebo-controlled, adequate sample size, timing of exposure in relation to ovulation verified by ultrasound and/or serum-hormone measurement. B= Assessed as moderate quality evidence. Placebo-or equivalent controlled, small or moderate sample size, ovulation timing by home urine test (21), serum hormone measurements (28), or repeated data from previous cycles (27). C= Assessed as low-quality evidence. No control group, small or moderate sample size, timing of ovulation verified by ultrasound and/or serum-hormone measurement (22, 24, 26) or menstrual cycle length (25).
Follicle size 12−14
Follicle size 15−17
Follicle size ≥18 mm
At LH surge start
Before LH peak
At LH peak
Brache et al, 2013
RCTs
98
(n = 98)LNG 14.6% vs placebo 4%, p > 0.05.
(n = 18) LNG 25% vs placebo 0%
(n = 24): LNG 14% vs placebo 10%
(n = 46): LNG 9% vs placebo 4%
A
Croxatto et al, 2004
RCT
110
(n = 36) LNG 94% vs placebo 61% (p > 0.05)
(n = 44) LNG 91% vs placebo 45%, p = 0.003
(n = 33) LNG 47% vs placebo 13% (p > 0.05)
A
Hapangama et al, 2001
RCT
24
(n = 24) LNG 42% vs placebo 8%, p-value> 0.05
B
Marions et al, 2002
Prospective cohort
12
(n = 12)LNG 83% vs placebo 17% (p < 0.05)
B
Marions et al, 2004
Descriptive
7
(n = 7) LNG 100% control cycle not reported
C
Natavio et al, 2018
Descriptive
24
(n = 24) LNG 79% no control group
C
Tirelli et al, 2008
Descriptive
8
(n = 7)100%
(n = 1)0%
C
Okewole et al, 2007
Descriptive
28
(n = 8) LNG 50%, no control
(n = 6) LNG 0%, no control
C
Treatment outcome: observed versus expected pregnancies
All studies compared observed number of pregnancies to a global estimate of expected pregnancy rates (29)
Before ovulation
After ovulation
Novikova et al, 2007
Descriptive
51
(n = 34) LNG 0 vs 4.2 expected
(n = 17) LNG 3 vs 3.5 expected
B
Noe et al, 2011
Descriptive
148
(n = 34) LNG 0 vs 16 expected(p < 0.001)
(n = 45) LNG 8 vs 8.7 expected (p = 1.00)
B
LH, luteinizing hormone; n, number; RCT, randomized controlled trial; LNG, levonorgestrel EC (All studies used oral administration of LNG 1.5 mg, or 0.75 mg given twice orally at a 12 hours interval, except for Brache et al. 2007 where a single vaginal dose of 0.75 mg was used). Significant changes are marked in bold.
a Inhibited or delayed ovulation for ≥5 days.
b All studies compared observed number of pregnancies to a global estimate of expected pregnancy rates (29)
c Grading of evidence:A= Assessed as high-quality evidence. Placebo-controlled, adequate sample size, timing of exposure in relation to ovulation verified by ultrasound and/or serum-hormone measurement.B= Assessed as moderate quality evidence. Placebo-or equivalent controlled, small or moderate sample size, ovulation timing by home urine test (21), serum hormone measurements (28), or repeated data from previous cycles (27).C= Assessed as low-quality evidence. No control group, small or moderate sample size, timing of ovulation verified by ultrasound and/or serum-hormone measurement (22, 24, 26) or menstrual cycle length (25).
The studies were of moderate quality. When Levonorgestrel EC was taken prior to ovulation, one study reported 0 pregnancies vs 4 expected, and the other 0 vs 16 expected (p < 0.001) [
]. Corresponding figures for both studies after intake of Levonorgestrel EC directly after ovulation were 3 recorded pregnancies vs 3.5 expected, and 8 vs 8.7 expected (p = 1.00).
Summarized data for all studies reporting on ovulation and conception included in the review are presented in Table 1.
3.3 Implantation
We identified 10 studies on endometrial receptivity measures after Levonorgestrel EC exposure. Two in vitro studies showed that treatment of 3-dimensional human endometrial tissue construct with 10μM levonorgestrel did not significantly alter human blastocyst attachment rate [
]. Five in vivo biomedical studies showed that Levonorgestrel EC administered either before, on the day of, or after LH surge resulted in no change in endometrial receptivity markers compared to controls [
Analysis of the endometrial transcriptome at the time of implantation in women receiving a single post-ovulatory dose of levonorgestrel or mifepristone.
In vivo assessment of the human sperm acrosome reaction and the expression of glycodelin-A in human endometrium after levonorgestrel-emergency contraceptive pill administration.
A single midcycle dose of levonorgestrel similar to emergency contraceptive does not alter the expression of the L-selectin ligand or molecular markers of endometrial receptivity.
Fertil Steril.2010; 94 (doi: 10.016/j.fertnstert.2009.09.013. Epub Nov 11): 1589-1594
] and the authors concluded that levonorgestrel did not affect implantation. Two in vivo biomedical studies showed that Levonorgestrel EC taken before the LH surge induced premature rise in serum and intrauterine concentrations of glycodelin-A in the peri-ovulatory phase compared to controls, which might inhibit sperm-oocyte interaction if ovulation were to occur [
], but a weaker expression of glycodelin-A on day 9 following the LH surge (LH +9) or no significant change when taken on or after the day of LH surge, indicating lack of effect on implantation [
Late follicular phase administration of levonorgestrel as an emergency contraceptive changes the secretory pattern of glycodelin in serum and endometrium during the luteal phase of the menstrual cycle.
]. Another in vivo study showed that vaginal Levonorgestrel EC exposure after the LH surge had no effect on 14 endometrial receptivity markers, but that oral exposure affected the expression of three biomarkers: a reduction in expression of progesterone receptors (PR-A and PR-B) in the glandular epithelium, and increased expression of leukemia inhibitory factor (LIF) [
]. Progesterone directly or indirectly regulates many of the endometrial cytokines involved in implantation. LIF is a cytokine whose mechanism of action in human implantation is unclear, most evidence supports that reduced, but not higher, levels, are associated with lower conception rates and higher risk of miscarriage [
All in vitro concentrations have been converted to micromoles(uM). Peak serum concentration of LNG after a single 0.75 mg or 1.5 mg oral dose are 0.02 μM and 0.039 μM respectively.
All in vitro concentrations have been converted to micromoles(uM). Peak serum concentration of LNG after a single 0.75 mg or 1.5 mg oral dose are 0.02 μM and 0.039 μM respectively.
Marions et al, 2002
In vivo biomedical study, control and treatment cycles in same subjects
6
Two doses of 0.75 mg LNG taken 12H apart on LH-2 D (estimated) and LH +2 D
LNG treatment produced no significant change in endometrial receptivity markers including integrin α4 and β3, cyclooxygenase-1 and -2, progesterone receptors, Dolichos biflorus agglutinin lectin binding and pinopodes in endometrial biopsy on LH +6 to 8 D.
B
Durand et al, 2005
In vivo biomedical study, control and treatment cycles in same subjects
30
Two doses of 0.75 mg LNG taken 12 H apart on:Group 1: LH -4 or -3 DGroup 2: LH dayGroup 3: LH +2 D
Endometrial biopsy on LH+9 D showed weaker endometrial glycodelin-A expression in Group 1 in the treatment cycle compared with the pre-treatment control cycle and when compared to Groups 2 and 3. No differences were found between control and treatment cycles in Groups 2 and 3.
A
Do Nascimento et al, 2007
In vivo biomedical study, double-blind randomised placebo-controlled
15(number contributing to endometrial study unclear)
LNG 1.5 mg 24H after intrauterine insemination
No change in endometrial glycodelin-A protein expression by immunostaining at 24 and 48H after LNG intake.
B
Lalitkumar et al, 2007
In vitro blastocyst-endometrial co-culture study, in vitro control
14 (LNG group) and 17 (control)
Progesterone and LNG at 10μM after 5−6 D of endometrial culture
LNG treatment did not significantly alter the human blastocyst attachment rate to the in vitro 3-dimensional endometrial construct (43%) compared to control (59%).
B
Meng et al, 2009
In vitro study, in vitro control
9 (LNG) and 12 (control)
10μM LNG in vitro treatment
LNG treatment produced no significant change in the expression of endometrial receptivity markers studied including ER-beta, PR-B, VEGF, integrin avb3 and MUC1 in 3-dimensional in vitro human endometrial tissue constructs.
B
Meng et al, 2009
In vivo biomedical study, non-randomised
9 (exposed to LNG-EC but got pregnant and resorting to induced abortion) and 9 (unexposed controls)
LNG 1.5mg on cycle day 16−23 for emergency contraception
Women with LNG exposure before pregnancy did not show significant change in the immunohistochemical expressions of ER-alpha, ER-beta, PR-B, PR total, androgen receptor or proliferation index Ki67 in the first-trimester decidua and chorionic villi.
B
Meng et al, 2010
In vivo biomedical study, control and treatment cycles in same subjects
8 (oral group) and 7 (vaginal group)
Oral LNG 0.75 mg x 4 doses at 24H intervals on LH +1 to +4 D, or vaginal LNG 1.5 mg single dose on LH +2 D
In the oral group, LNG treatment resulted in reduced immunoreactivity of PR-A and PR-B in glandular epithelium, and increased stromal immunoreactivity and mRNA expression of LIF compared with control cycle. Vaginal LNG treatment did not cause any significant change in endometrial markers.
B
Palomino et al, 2010
In vivo biomedical study, randomised single-blind
14 (oral group), 13 (vaginal group) and 11 (control group with no treatment)
LNG 1.5 mg orally or vaginally on the day of LH surge
LNG treatment did not alter expression of receptive markers including PR, L-selectin ligand, αvβ3 integrin and glycodelin-A in endometrial biopsy taken on LH+2 and +7 D.
B
Durand et al, 2010
In vivo biomedical study, control and treatment cycles in same subjects
30
Two doses of 0.75 mg LNG taken 12H apart on LH-2 D
Serum and uterine flushing glycodelin-A on LH+1 D showed significant premature rise in the treatment cycle compared with the pre-treatment control cycle.
A
Barrios-Hernández AE et al, 2020
In vivo biomedical study, control and treatment cycles in same subjects
5
LNG 1.5 mg on day of ovulation (follicle rupture)
LNG did not alter the endometrial transcriptome studied on ovulation +6 D (the implantation window) in the treatment cycle compared with the pre-treatment cycles.
B
LH, luteinizing hormone; LNG, levonorgestrel; n, number.
Quality of evidence
A= Assessed as high quality of evidence. Placebo-or equivalent controlled, sample size≥20, timing of exposure in relation to ovulation verified by ultrasound and/or serum-hormone measurement.
B = Assessed as moderate quality of evidence. Placebo-or equivalent controlled, sample size<20, timing of exposure in relation to ovulation verified by ultrasound and/or serum-hormone measurement.
C = Assessed as low quality of evidence low. Uncontrolled, timing of exposure in relation to ovulation uncertain
a All in vitro concentrations have been converted to micromoles(uM). Peak serum concentration of LNG after a single 0.75 mg or 1.5 mg oral dose are 0.02 μM and 0.039 μM respectively.
Expressions of candidate molecules in the human fallopian tube and chorionic villi of tubal pregnancy exposed to levonorgestrel emergency contraception.
Expressions of candidate molecules in the human fallopian tube and chorionic villi of tubal pregnancy exposed to levonorgestrel emergency contraception.
]. Two studies found a dose-dependent reduction in ciliary beat frequency in human fallopian tubes exposed to levonorgestrel compared to controls. Reduced motility was only seen at supra-pharmacological concentrations [
]. One study found that muscular contractions in the ampullo-isthmus area were reduced in a dose-dependent way after LNG exposure compared to controls (p < 0.05) [
]. The fifth study found that the expression of transient receptor potential vanilloid (TRPV) 4 channels, which regulate fallopian tube motility, were downregulated in women with ectopic pregnancy compared to women without ectopic pregnancy. TRPV4 channels were further downregulated among women with ectopic pregnancy who had taken Levonorgestrel EC (p < 0.05) [
]. Adjusted odds ratios in these studies were 10.5 (95% CI 2.21−46.56), 4.75 (95% CI 3.79−5.96), and 9.34 (95% CI: 3.9−16.0), respectively.
Data from all studies reporting on measures of ectopic pregnancy, and measures of fallopian tube motility, included in the review are summarised in Table 3.
Table 3Summary of original research studies reporting on the effects of LNG-emergency contraceptive on incidence of ectopic pregnancy or fallopian tube motility
All in vitro concentrations have been converted to micromoles(uM). Peak serum concentration of LNG after a single 0.75 mg or 1.5 mg oral dose are 0.02 μM and 0.039 μM respectively.
Assouni et al 2018
Case control study among 88 women with EP and 176 with IUP in Cameroon
264
Multivariate logistic regression analysis of association between EP and potential risk factors
Current use of LNG-EC as contraception AOR 10.15 (95% CI 2.21−46.56).
C
Shurie et al 2018
Case control study among 97 women with EP and 237 with IUP in Kenya
316
Multivariate logistic regression analysis of association between EP and potential risk factors
Use of LNG-EC in the cycle of conception AOR 9.34 (95% CI, 4.46−19.56).
C
Li et al 2015
Case control study among 2411 women with EP and 2416 with IUP in China
4827
Multivariate logistic regression analysis of association between EP and potential risk factors
Current use of LNG-EC as contraception AOR 4.75 (95% CI 3.79−5.96).
B
Li et al 2019
In vitro biomedical study on cultured fallopian tube epithelium explants and controls
58
Fallopian tube extract from 34 women with EP (of which 12 has taken LNG-EC and 22 had not) and 24 controls.
There were significant (p < 0.05) lower protein expressions of TRPV4 (a calcium channel involved in tubal motility) in women with EP and even lower (p < 0.05) among women with EP exposed to LNG-EC compared to women without EP.
A
Li et al 2018
In vitro biomedical study on cultured fallopian tube epithelium explants and controls
3+3
Tubal epithelium exposed to LNG concentrations ranging from 0.01 μM to 10 μM
CBF dose-dependent decrease. No significant CBF decrease at pharmacological concentrations. No change in levels of fallopian tubal epithelial cell receptivity markers, including LIF, STAT3, IGFBP1, ITGB3, MUC1, and ACVR1B after LNG exposure.
B
Zhao et at 2015
In vitro biomedical study on cultured fallopian tube epithelium explants and controls
10 +10
Tubal epithelium from the proliferative and secretory phase exposed 5. μM and 0.5μM LNG, both concentrations are suprapharmacological.
CBF decreased in the secretory and proliferative cycle phase after suprapharmacological exposure to 5.0 μM of LNG. At 0.5 μM CBF decreased only in the ampulla and only during the secretory phase. Cilia morphology was unchanged by LNG at any concentration.
A
Huang et al 2013
In vivo biomedical study on fallopian tube extracts
55
27 women exposed to LNG-EC before the EC and 28 women unexposed.
No difference in mRNA or protein expressions of ER-alpha, PR, LIF, VEGF, iNOS and CB1 in fallopian tubes of women exposed to LNG-EC compared to unexposed women.
A
Wånggren et al 2008
In vitro biomedical study on cultured fallopian tube epithelium explants and controls
22+22
Tubal epithelium exposed to supra- (0.2 μM) and pharmacological (0.02 μM) concentrations of LNG.
CBF decreased in the ampullo-isthmus area in a dose-dependent way after exposure to both suprapharmacologic al and pharmacological concentrations of LNG. The decrease at a pharmacological dose was uncertain.
A= Assessed as high quality of evidence. Placebo-or equivalent controlled, sample size≥20, timing of exposure in relation to ovulation verified, interventions clearly described.
B = Assessed as moderate quality of evidence. Placebo-or equivalent controlled or large epidemiological study, sample size<20, timing of exposure in relation to ovulation verified or interventions clearly described.
C = Assessed as low quality of evidence. Uncontrolled, timing of exposure in relation to ovulation uncertain, or small epidemiological study with inadequate power as reflected by large confidence intervals.
a All in vitro concentrations have been converted to micromoles(uM). Peak serum concentration of LNG after a single 0.75 mg or 1.5 mg oral dose are 0.02 μM and 0.039 μM respectively.
Failure of the emergency contraceptive levonorgestrel and the risk of adverse effects in pregnancy and on fetal development: an observational cohort study.
]. In one cohort study, researchers compared pregnancy outcomes between 332 women exposed to Levonorgestrel EC in early pregnancy and 332 unexposed women [
]. There was no significant difference in rate of miscarriage (10.3 vs 8.6%, p = 0.47) or congenital malformations (1.5 vs 1.3%, RR 1.1, 95% confidence interval 0.28−4.37). A study from the same setting subsequently compared rate of congenital malformations between 195 exposed and 214 unexposed women with similar findings (2.1 vs 1.4%, p > 0.05). Two-year follow-up found no differences in height, weight, head circumference or intelligence between exposed and unexposed participants [
]. A third study reported on similar outcomes after exposure to Levonorgestrel EC (n = 25) compared to a non-exposed group (n = 80). There were no significant differences in the rate of miscarriage (p > 0.05) or congenital malformations, but the study was considered underpowered to assess these outcomes [
Failure of the emergency contraceptive levonorgestrel and the risk of adverse effects in pregnancy and on fetal development: an observational cohort study.
We identified no studies reporting on increased use of assisted reproduction treatment or reported infertility after Levonorgestrel EC exposure, time to conception in subsequent cycles, or fertility within 12 months of exposure.
We identified three studies that reported on menstrual bleeding patterns in the cycle following Levonorgestrel EC intake (second cycle), but no studies on ovulation in this cycle. According to all studies, Levonorgestrel EC was, as expected, associated with significant change in menstrual pattern in the exposed cycle but no change in the second cycle. A prospective observational study of 232 women with normal baseline menstrual cycles (27–29 days) found that the second cycle was not significantly different from the pre-exposure cycle [
]. Another study among 69 women found cycle length was normalized in the second cycle irrespective of timing of Levonorgestrel EC intake in relation to ovulation in the exposed cycle [
This review confirms that Levonorgestrel EC acts by postponing or inhibiting ovulation and that there is little to no post-ovulatory action. We found no evidence to support that exposure to Levonorgestrel EC affects fetal development, miscarriage, stillbirth, or subsequent menstruations.
4.2 Interpretation
It has been argued that Levonorgestrel EC inhibits implantation since the rate of ovulatory inhibition is too small to account for the overall effectiveness of the treatment and that Levonorgestrel EC may therefore be a form of “emergency abortion”. Original research articles in support of a postovulatory effect are lacking [
]. The synthesized data in this review finds no evidence that Levonorgestrel EC taken after ovulation prevents implantation. If Levonorgestrel EC does not inhibit ovulation, the resulting conception rates are similar to those expected after unprotected intercourse in the high-fertile period. This is supported by a Cochrane meta-analysis in which women with further acts of unprotected intercourse in the same cycle after use of hormonal EC (including levonorgestrel) had significantly higher risk of pregnancy compared to those without further unprotected intercourse. This implies that when ovulation is postponed, fertility is still preserved in the later part of the same cycle after EC use [
Some studies have suggested that the efficacy of Levonorgestrel EC has been overestimated in clinical studies and that its real-life efficacy is between 50 and 75% [
]. According to these researchers, this overestimation has lent unfounded support to the theories of “missing mechanism of action” of Levonorgestrel EC [
]. Another explanation for the discrepancy between levonorgestrel´s inhibitory effect on ovulation and its total efficacy in preventing pregnancy, is that the standard definition of inhibited ovulation underestimates Levonorgestrel EC´s inhibitory effect. A 5-day delay is the standard definition of inhibited ovulation, based on the concept of 5-day sperm viability. Depending on when intercourse occurs however, a 2-to-3-day delay could be sufficient to avoid fertilization. Most reviewed studies indicate a statistically non-significant difference between placebo and Levonorgestrel EC when given close to the LH surge or at follicle size ≥18mm. Most studies were small, and these differences may have been significant in a larger study. Levonorgestrel EC could therefore have a window of action that extends to, or even includes, the LH surge that explains that its overall effect in practice exceeds the rate of inhibited ovulation reported in these studies.
In our review, 9 of 10 studies supported that although levonorgestrel sometimes affects endometrial receptivity if taken before the LH surge, it has no effect on receptivity or blastocyst attachment after ovulation. The effect seen after pre-ovulatory intake is likely secondary to the lack of luteal activity following inhibited ovulation. One study found that 3 of 14 receptivity biomarkers were reduced after LNG administration after the LH surge with oral but not vaginal administration. Overall, the evidence is overwhelmingly in support of a lack of postovulatory implantation effect. Furthermore, there is no evidence to demonstrate that changes in endometrial receptivity biomarkers translate into clinical effects on embryo implantation. In vivo animal studies have convincingly shown that Levonorgestrel EC has no postfertilization effect and that conception rates are not different from those after placebo [
With one exception, in vitro studies included in our review support that Levonorgestrel EC, at doses equivalent to those used for EC, does not affect the receptivity in the fallopian tubes in a way to increase the risk of extrauterine implantation. Several in vitro studies indicate that exposure at concentrations tenfold those seen in serum after oral intake of Levonorgestrel EC, reduces fallopian tube motility. According to all these studies the effect is dose-dependent and ceases or is minimal at pharmacological concentration. We found one vivo study that reported that ion channels involved in fallopian tube motility were downregulated in women with ectopic pregnancy exposed to Levonorgestrel EC. Channels were downregulated also in women with ectopic pregnancy compared to those without, so the role of Levonorgestrel EC, as well as how the finding translates to clinical tubal motility, was uncertain. We found 3 epidemiological studies that found an increased risk of ectopic pregnancy after Levonorgestrel EC use of which only one was large enough to address the proposed outcomes. Other known potential confounders for ectopic pregnancy were adjusted for but residual confounding may explain the association seen. Post-marketing surveillance data have found no association between ectopic pregnancy and Levonorgestrel EC use [
]. Similarly, a systematic review in 2011 that collated safety data, including ectopic pregnancy, from 25 studies on Levonorgestrel EC found a cumulative incidence of ectopic pregnancy of 1% after EC use which is not higher than expected baseline rates [
]. The in vitro and epidemiological results synthesized in the review are conflicting, however, any increased relative risk of ectopic pregnancy after Levonorgestrel EC use must be viewed in the context of the small risk in absolute terms of the combined occurrence of contraceptive failure, conception, and baseline risk of ectopic pregnancy.
We found no increased risk of miscarriage, fetal abnormality or child development problems in pregnancies occurring after Levonorgestrel EC exposure which is supported by previous systematic reviews on the safety of Levonorgestrel EC use [
]. We found no indication that Levonorgestrel EC disrupts the return of a normal menstrual cycle in the cycle following exposure. Although we found no studies on long-term fertility after Levonorgestrel EC use, there is substantial evidence that other levonorgestrel-only contraceptives do not impair future fertility, even after prolonged use. The half-life of levonorgestrel after discontinuation of the levonorgestrel-implant is estimated at 42 +/- 16 hours, with only trace amounts remaining after 96 hours and clearance is not influenced by duration of use [
]. Against this background repeated use of Levonorgestrel EC is unlikely to affect future fertility. Three systematic reviews on future fertility after contraceptive use have all concluded that Levonorgestrel containing contraceptives, irrespective of parity, gravidity or duration of use, have no detrimental effect on time to pregnancy or fecundity compared to non-hormonal methods and that conception rates are similar to those expected in the general population [
Levonorgestrel EC is a safe method that reduces the risk of an unplanned pregnancy by inhibiting ovulation. Our review supports that, women can use this medication without risking adverse effects in a pregnancy exposed to Levonorgestrel EC effect and that concerns about "abortifacient" effects are unfounded. While there is no definitive evidence regarding the impact of Levonorgestrel EC on future fertility, the body of evidence for other Levonorgestrel contraceptives suggests there should be minimal effect.
4.3 Strengths and limitations
This review encompassed a wide range of outcomes relating to pregnancy viability and future fertility after exposure to Levonorgestrel EC. The search was systematic to retrieve all available data, data were screened by two independent researchers using standardized criteria to assess quality of evidence. Results are presented in descriptive form so as not to dissimulate the heterogeneity of study designs and outcome measures. Many of the clinical studies were small which lends uncertainty to the results. A further limitation to the data relates to the uncertainty of time of ovulation. Time of ovulation in relation to exposure was central to most outcomes reviewed in the study. This measure is however challenged by the fact that it is difficult to pinpoint the exact timing of LH surge, follicle size, and serum hormone levels in relation to ovulation. We included an assessment of the reliability of this measure in our quality assessment.
5. Conclusion
Based on the current available evidence, Levonorgestrel EC acts by inhibiting or postponing ovulation, but has no significant post-ovulatory action. There is no evidence that exposure to Levonorgestrel EC affects fetal development, miscarriage, stillbirth, or subsequent menstruations.
Author contributions
ME and KGD designed the study. ME designed and performed the systematic search. ME and RHWL performed the review of selected studies and analysis of data. All authors contributed to the finalization of the manuscript.
Funding
The research was funded in part by the Foundation Consumer Healthcare.
Role of the funding source
The funder played no role in the study design, data collection, analysis, or presentation of the findings in this article.
Declaration of Competing Interest
ME received a partial salary for the research involved in performing the systematic review from Foundation Consumer Healthcare. She has no other conflicts of interest to report. KGD has served as ad hoc invited speaker or on advisory meetings for Gedeon Richter, Bayer AG, MSD, Exelgyn, Exeltis, Azanta, HRA Pharma, and Ferring. RHWL has received research support from HRA Pharma and Ferring and served as an ad hoc invited speaker for Ferring.
A prospective, open-label, single arm, multicentre study to evaluate efficacy, safety and acceptability of pericoital oral contraception using levonorgestrel 1.5 mg.
PATH, the World Health Organization, and the United Nations Population Fund. Essential Medicines for Reproductive Health: Guiding Principles for Their Inclusion on National Medicines Lists. Seattle: PATH; 2006.
Key statistics from the National Survey of Family Growth . Centers for disease control and prevention. Atlanta: National Center for Health Statistics; 2017.
Use of non-emergency contraceptive pills and concoctions as emergency contraception among Nigerian University students: results of a qualitative study.
Ulipristal acetate prevents ovulation more effectively than levonorgestrel: analysis of pooled data from three randomized trials of emergency contraception regimens.
Effect of a single vaginal administration of levonorgestrel in Carraguard (R) gel on the ovulatory process: a potential candidate for "dual protection" emergency contraception.
Pituitary-ovarian function following the standard levonorgestrel emergency contraceptive dose or a single 0.75-mg dose given on the days preceding ovulation.
Analysis of the endometrial transcriptome at the time of implantation in women receiving a single post-ovulatory dose of levonorgestrel or mifepristone.
In vivo assessment of the human sperm acrosome reaction and the expression of glycodelin-A in human endometrium after levonorgestrel-emergency contraceptive pill administration.
A single midcycle dose of levonorgestrel similar to emergency contraceptive does not alter the expression of the L-selectin ligand or molecular markers of endometrial receptivity.
Fertil Steril.2010; 94 (doi: 10.016/j.fertnstert.2009.09.013. Epub Nov 11): 1589-1594
Late follicular phase administration of levonorgestrel as an emergency contraceptive changes the secretory pattern of glycodelin in serum and endometrium during the luteal phase of the menstrual cycle.
Expressions of candidate molecules in the human fallopian tube and chorionic villi of tubal pregnancy exposed to levonorgestrel emergency contraception.
Failure of the emergency contraceptive levonorgestrel and the risk of adverse effects in pregnancy and on fetal development: an observational cohort study.
00006-3&id=gr1.jpg){kind=link}