Relationship Between Etonogestrel Concentrations and Bleeding Patterns in Contraceptive Implant Users

Abstract

Objectives

To estimate whether serum etonogestrel concentrations influence bleeding patterns and related side effects in contraceptive implant users.

Methods

We conducted a prospective cross-sectional study with healthy, reproductive-aged women using etonogestrel implants for 12–36 months. Participants completed a brief questionnaire to assess their current bleeding pattern and any experience of abnormal bleeding with the implant. We then measured serum etonogestrel concentrations. We also reviewed the charts of participants to determine if a prescription for oral contraceptive pills was ever given for treatment of implant-related bothersome bleeding. We performed multivariable logistic regression to test for associations between serum etonogestrel concentrations and both bleeding patterns and related side effects.

Results

We enrolled 350 women and 59.4% reported having experienced abnormal bleeding with the contraceptive implant. Only 14.9% of participants reported amenorrhea and 37.7% reported monthly periods. Among participants with reviewable medical records (n=253), roughly 20% had received a prescription for oral contraceptive pills during implant use. Increasing serum etonogestrel concentrations were significantly associated with increasing odds of reporting abnormal bleeding (aOR 1.005, p=0.015) and increasing odds of having received an oral contraceptive pill prescription (aOR 1.008, p=0.002). For every 100pg/mL increase in serum etonogestrel concentration, contraceptive implant users in this study had 1.6 times the odds of reporting abnormal bleeding and 2.3 times the odds of having received a prescription as treatment for bothersome bleeding.

Conclusion

We found both objective and subjective evidence that higher levels of progestin from the contraceptive implant were associated with bleeding side effects experienced by women in this study. Pharmacologic variation may influence the side effects women experience with a variety of hormonal contraceptive methods, in turn affecting patient satisfaction and discontinuation rates.

Précis

Higher levels of circulating progestin are associated with higher rates of abnormal bleeding and increased need for interventions for bothersome bleeding among contraceptive implant users.

Introduction

Use of the etonogestrel contraceptive implant (Nexplanon^®, formerly Implanon^®, Merck & Co., Whitehouse Station NJ) continues to increase in the United States, especially among young adults and adolescents (1, 2). The etonogestrel implant has a relatively high continuation rate (65–91% at one year of use); however, 11.3% of women ultimately discontinue this method due to bothersome bleeding irregularities (3–5). Prolonged periods of bleeding and frequent irregular bleeding were the most common bleeding patterns associated with discontinuation (5). In the St. Louis-based contraceptive CHOICE project, the majority of women who discontinued the contraceptive implant within the first 6 months of use (53%) reported “Irregular or Frequent bleeding” as a reason (6). Though many treatments have been studied to treat bothersome bleeding with the etonogestrel implant (e.g. oral contraceptive pills, non-steroidal anti-inflammatory drugs, tamoxifen), these treatments often provide only short-term benefits with resumption of abnormal bleeding after treatment cessation (7–9).

Despite ample data on the prevalence of abnormal bleeding patterns with the contraceptive implant, little published data exist on how individual patient characteristics influence a woman’s risk for experiencing bleeding-associated side effects. In addition to the wide variability in bleeding profiles, several studies have demonstrated wide variability in interindividual serum etonogestrel concentrations among women using the etonogestrel implant for identical periods of time (10, 11). The two largest published cohorts of contraceptive implant users found a range of serum etonogestrel concentrations of 55.8–802.6pg/mL, a greater than 14-fold difference between the highest and lowest values (10, 11). We do not know the extent to which variability in serum drug levels, for the implant and other hormonal contraceptive methods, contributes to the variability in bleeding side effects (12, 13).

Given that bleeding-related side effects are some of the most common reasons for contraceptive method discontinuation, there is an urgent need to understand what factors may contribute to an individual woman’s risk for these side effects, including potential pharmacologic factors (3, 4, 6). As such, we aimed to evaluate the relationship between serum etonogestrel concentrations and both bleeding patterns and related side effects in contraceptive implant users. Etonogestrel contraceptive implant users are ideal candidates for pharmacokinetic variability studies as they achieve relative steady-state serum concentrations due to the steady drug release nature of the implant. This avoids any confounding stemming from protocol adherence (4, 14). We hypothesized that increased serum etonogestrel concentrations would be associated with increased odds of experiencing abnormal or bothersome bleeding with the contraceptive implant.

Methods

The underlying methodology for this study has previously been published, including all inclusion and exclusion criteria (10). Participants were English or Spanish speaking reproductive age women (18–45 years old) with an etonogestrel contraceptive implant in place for at least 12 and no more than 36 months. Women had to have an implant in place for at least 12 months in order to be in the relative steady-state period of the implant’s pharmacokinetics (14). From the available pharmacokinetic data, the amount of etonogestrel released from the implant decreases by only 15% on average between 1 and 2 years of use (40ug per day versus 34ug per day), as compared to a decrease of up to 33% between 4 weeks (60–70ug per day) and 1 year of use (14). We chose 36 months as our upper enrollment limit for duration of implant use as this is the current Food and Drug Administration approved duration of use (4). We excluded women using any medications or supplements that could affect serum etonogestrel levels through inhibition or induction of cytochrome P450 (CYP) enzymes (specifically CYP3A4) (15). We also excluded women who reported any medical conditions that could affect baseline liver function (e.g. hepatitis, cirrhosis) or a measured BMI less than 18.5kg/m², as low BMI has been associated with abnormal metabolism. The hypothesis tested in this analysis was a pre-specified outcome of a broader pharmacogenomic study (10). The protocol was approved by the Colorado Multiple Institutional Review Board and all participants gave written informed consent before study initiation. This study was also registered on ClinicalTrials.gov (NCT03092037). We recruited participants through community advertising and contraceptive clinics at the University of Colorado Anschutz Medical Campus and compensated participants with a $20 gift card for their participation.

For eligible participants, we determined the length of implant use by participant self-report and confirmed presence of the implant by physical exam. We excluded women who could not remember when they had the implant placed, unless there was documentation of the date of insertion in their medical record. Participants completed a questionnaire to obtain self-reported demographics and side effect information (Appendix 1, available online at http://links.lww.com/xxx). To specifically assess subjective bleeding patterns and side effects, we asked participants if they had experienced “abnormal bleeding” (yes/no) while using their contraceptive implant. We also assessed if participants experienced monthly periods (yes/no) at the time of enrollment, and if not, roughly how many days of bleeding or spotting they had over the last 90 day (3 month) period prior to enrollment. When defining bleeding patterns, we designated participants who reported 0 days of bleeding or spotting over a 90 day period as experiencing amenorrhea. For participants who were recruited during a routine visit with their healthcare provider at our institution, we also reviewed their medical record to determine if they had received a prescription for combined oral contraceptive pills for treatment of bothersome breakthrough bleeding at any time during use of their contraceptive implant. We specifically evaluated combined oral contraceptive pill prescription as this is the standard treatment for bothersome breakthrough bleeding with the contraceptive implant at our recruitment sites.

We performed serum etonogestrel concentration analysis using the previously published methodology that utilized a previously validated liquid chromatography-mass-spectrometry (LCMS) assay protocol (10, 16). All LCMS procedures were performed at the Biomarkers Core Laboratory of the Irving Institute of Clinical and Translational Research at Columbia University Medical Center (New York City, NY).

We used IBM SPSS™ version 25 statistical software for all statistical analyses. We performed descriptive frequencies for reported abnormal bleeding and bleeding patterns. We also performed multivariable logistic regression to identify predictors for abnormal bleeding, for the bleeding patterns of monthly periods and amenorrhea, and for the provision of oral contraceptive pills for treatment of bothersome bleeding. We chose the pertinent patient characteristics of age, months of implant use, BMI, race, ethnicity, and serum etonogestrel concentration as independent variables for our logistic regression analyses. We utilized a backward conditional approach to create our logistic regression model for our selected outcomes of interest. The sample size was determined by the primary pharmacogenomic study size (10).

Results

We recruited 350 participants over the course of 15 months (March 2016 to May 2017). Table 1 shows pertinent patient characteristics. The median age of participants was 22.5 years with a median duration of etonogestrel implant use of 26.0 months. The most frequent self-reported race was White or Caucasian (46.6%) and 51.4% of participants reported Hispanic or Latina ethnicity. Amongst the 77 participants who responded with “No response or Unknown” in relation to self-identified race, all identified their ethnicity as “Hispanic or Latina.” The median serum etonogestrel concentration was 137.4pg/mL (range 55.8–695.1) with an interquartile range of 63.5pg/mL.

Table 1:

Participant characteristics and demographics (N=350)

	Median (Range)
Age (years)	22.5 (18.0 – 39.1)
Months of implant use	26.0 (12.0 – 36.0)
BMI (kg/m2)	25.7 (18.5 – 52.0)
Serum etonogestrel concentration (pg/mL)	137.4 (55.8 – 695.1)
	n (%)
Race
White or Caucasian	163 (46.6)
Black or African American	40 (11.4)
Asian or Pacific Islander	19 (5.4)
Native American or Alaskan	7 (2.0)
More than one	44 (12.6)
No response or Unknown	77 (22.0)
Ethnicity
Hispanic or Latina	180 (51.4)
Non-Hispanic	170 (48.6)

The majority of participants (59.4%) reported experiencing abnormal bleeding and 37.7% of participants responded that they currently have a monthly period. Table 2 demonstrates the full breakdown of reported bleeding patterns. Using the definition of amenorrhea as outlined in the methods, 52 of the 205 participants without a monthly period were designated as having amenorrhea (25.4%). Overall, a large majority of participants without monthly periods reported less than 16 days of bleeding or spotting over a 90 day period (71.7%), with only 10.2% of participants reporting bleeding in excess of 45 days.

Table 2:

Frequencies of bleeding-related side effects and bleeding patterns among contraceptive implant users.

	Yes n (%)
Abnormal bleeding*	208 (59.4)
Current monthly period*	132 (37.7)
Amenorrhea*	52 (14.9)
Oral contraceptive pill prescription†	53 (20.9)
Days of bleeding±	n (%)	Cumulative %
0	52 (25.4)
1 – 8	60 (29.2)	54.6
9 – 15	35 (17.1)	71.7
16 – 30	27 (13.2)	84.9
31 – 45	10 (4.9)	89.8
>45	21 (10.2)	100.0
Bleeding patterns of participants prescribed oral contraceptive pills§	n (%)
Current monthly period	26 (49.1)
Amenorrhea	4 (7.6)
1 – 15 days of bleeding	5 (9.4)
16 – 30 days of bleeding	5 (9.4)
>30 days of bleeding	13 (24.5)

^*Among all 350 participants. These categories were not mutually exclusive: 91 participants who reported a “current monthly period” also reported ever having “abnormal bleeding” and 14 participants who had “amenorrhea” at the time of enrollment reported ever having “abnormal bleeding”

^†Among 253 participants recruited at routine healthcare visits

^±Among 205 participants who reported not having a current monthly period. Days of bleeding data was missing for 13 participants.

^§Among 53 participants

Almost three-fourths of participants were recruited during routine healthcare visits (72.3% [253/350]). Among this group, 20.9% (53/253) had received a prescription for combined oral contraceptive pills for management of bothersome bleeding during the time they had an etonogestrel contraceptive implant in place. No participants were prescribed oral contraceptive pills containing progestins that are metabolized into or directly related to etonogestrel (e.g. desogestrel).

For the multivariable logistic regression analysis, we included the participant characteristics and demographics of age, months of implant use, BMI, race, ethnicity, and serum etonogestrel concentration as our variables of interest based on availability and plausible associations with bleeding profiles. The crude odds ratios (ORs) for all variables of interest and our bleeding side effect outcomes are shown in Table 3. The final multivariable logistic regression models and adjusted odds ratios (aORs) are presented in Table 4. Overall, serum etonogestrel concentrations were significantly associated with two outcomes. For every 1pg/mL increase in serum etonogestrel concentration, participants had 1.005 times the odds of having experienced abnormal bleeding (p=0.015) and 1.008 times the odds of having received a prescription for oral contraceptive pills for bothersome bleeding (p=0.002). No other variables remained in the logistic regression model for abnormal bleeding.

Table 3:

Univariate logistic regression analysis for crude odds ratios between all patient characteristics and demographics and bleeding patterns and outcomes of interest

	Abnormal bleeding (yes/no)		Current monthly period (yes/no)		Amenorrhea (yes/no)		Oral contraceptive pill prescription (yes/no)
	OR	95% CI	OR	95% CI	OR	95% CI	OR	95% CI
Age	0.97	0.91 – 1.03	1.02	0.95 – 1.08	1.00	0.91 – 1.09	1.03	0.88 – 1.22
Duration of implant use	1.0	0.97 – 1.02	0.99	0.97 – 1.01	0.99	0.96 – 1.02	0.94†	0.91 – 0.97
BMI	0.98	0.94 – 1.01	0.98	0.95 – 1.02	1.01	0.97 – 1.07	0.92†	0.89 – 098
Serum etonogestrel concentration	1.005†	1.001 – 1.008	1.000	0.997 – 1.003	1.00	0.99 – 1.00	1.011†	1.006 – 1.016
Race*
White or Caucasian	0.87	0.57 – 1.34	0.73	0.47 – 1.13	1.39	0.77 – 2.52	0.41†	0.20 – 0.80
Black or African American	1.16	0.59 – 2.28	2.22†	1.14 – 4.32	0.27	0.06 – 1.15	4.77†	2.11 – 10.80
Asian or Pacific Islander	1.51	0.56 – 4.07	0.96	0.37 – 2.51	3.25†	1.15 – 9.22	2.50	0.78 – 7.99
Native American or Alaskan	1.72	0.33 – 9.01	0.66	0.13 – 3.43	0.91	0.11 – 7.73	1.53	0.29 – 8.11
More than one	0.88	0.47 – 1.68	0.84	0.43 – 1.62	0.67	0.25 – 1.79	0.64	0.23 – 1.75
No response or Unknown	1.02	0.61 – 1.70	1.15	0.68 – 1.93	0.86	0.41 – 1.80	0.92	0.47 – 1.83
Ethnicity
Hispanic or Latina	0.91	0.59 – 1.40	1.01	0.65 – 1.55	0.88	0.49 – 1.59	0.56	0.30 – 1.03

^*Each race category was dummy coded for comparison to all other participants

^†Statistically significant using p-value cut-off <0.05

Table 4:

Final model of multivariable logistic regression analysis using a backward stepwise approach.

Side-effect/Bleeding pattern	Variables in final model	aOR	95% CI	p-value
Abnormal bleeding (yes/no)	Serum etonogestrel concentration (per 100pg/mL)	1.57	1.09, 2.27	0.015
Current monthly period (yes/no)	Black or African American race vs all other races	2.22	1.14, 4.32	0.019
Amenorrhea (yes/no)	Asian or Pacific Islander race vs all other races	3.25	1.15, 9.22	0.027
Oral contraceptive pill prescription (yes/no)	Duration of implant use*	0.95	0.91, 0.99	0.007
	Serum etonogestrel concentration (per 100pg/mL)	2.28	1.37, 3.80	0.002
	Black or African American race vs all other races	4.55	1.84, 11.26	0.001

^*Per month beyond the first 12 months of use

Two other variables were significantly associated with receiving a prescription for oral contraceptive pills, but with competing effects. For every 1 month of implant use past 12 months, participants had 0.95 times the odds of having received a prescription for oral contraceptive pills, but participants who self-reported as Black or African American had 4.5 times the odds of having been prescribed oral contraceptive pills as participants of any other race or ethnicity (Table 3). Black or African American race was also the only variable significantly associated with reporting a monthly period (aOR 2.22, 95% CI 1.14, 4.32). After accounting for oral contraceptive pill prescription, the association between Black or African American race and reporting a monthly period remained (aOR 2.33, 95% CI 1.02, 5.31). For amenorrhea, only Asian or Pacific Islander race remained as a significant predictor in the regression model (aOR 3.25, 95% CI 1.15, 9.22).

Discussion

Though having a monthly period or amenorrhea were not associated with serum etonogestrel concentrations, increases in serum etonogestrel concentration were associated with both increased subjective reporting of abnormal bleeding and increased objective prescription of combined oral contraceptive pills for bothersome bleeding. To put these associations into more clinically relevant terms, for every 100pg/mL increase in serum etonogestrel concentration, a woman had 1.6 times the odds of reporting having experienced abnormal bleeding and 2.3 times the odds of having received a prescription for oral contraceptive pills for management of bothersome bleeding. At our clinical setting, first-line treatment for bothersome bleeding with the etonogestrel implant is prescription of oral contraceptive pills. Thus, these findings support a clinically significant association between serum etonogestrel concentrations and bothersome bleeding side effects, as patients with higher serum etonogestrel concentrations were more likely to have received a medical intervention. Though higher BMI and longer duration of implant use are associated with small decreases in serum etonogestrel concentrations, we did not find that these two patient characteristics were associated with bothersome bleeding side effects (17). Given that discontinuation of the contraceptive implant and other hormonal contraceptive methods are often due to bothersome bleeding, understanding the role and sources of pharmacokinetic variability may lead to more individualized interventions (10).

In this large cohort of etonogestrel contraceptive implant users, we found similar bleeding patterns as demonstrated in prior studies (5). The percentage of participants in our study experiencing amenorrhea (15.4%) was somewhat lower than the rate found by Mansour et al. (22.2%), but that analysis included six studies conducted in Southeast Asian countries (5). We found that participants who self-reported their race as Asian or Pacific Islander had over 3 times increased odds of reporting amenorrhea as compared to all other participants, which could account for some of this discrepancy.

The only other self-reported race that was associated with a bleeding pattern was Black or African American; these participants had over 2 times the odds of reporting current monthly periods. However, it remains unclear from our findings why participants who self-reported as Black or African American had over 4 times the odds of having received a prescription for oral contraceptive pills. Since we did not assess the temporality of abnormal bleeding with our survey, it may be that these participants had bothersome bleeding patterns earlier in the use of the contraceptive implant, which then may have resolved for some of these women after prescription of oral contraceptive pills, or were more likely to report a desire for treatment of bothersome bleeding. Given this unresolved potential contradiction, research on bleeding patterns with other hormonal contraceptive methods may help elucidate the pertinence of our findings.

In this study were able to recruit a large and diverse cohort. Within this cohort, we were able to detect significant associations between our outcomes of interest and serum etonogestrel concentrations. Another strength of the study was recruiting participants using a medication with a controlled-release rate and steady-state pharmacokinetics, which allowed us to efficiently capture each individual’s pharmacokinetic profile with only a single etonogestrel measurement during the relative steady-state period between 12 and 36 months (4, 14). Our study population was also relatively young with a median age of 22.5 years, which is representative of the general population of etonogestrel implant users (4). Thus, the prevalence of bothersome bleeding and overall bleeding patterns of our study cohort are more likely to be representative of the general population of contraceptive implant users that continue past the first year of use.

The primary limitation of our study was that we used only a brief questionnaire to assess side effects and bleeding profiles. We designed this questionnaire to be short to reduce participant fatigue but our data lack some of the complexity in bleeding profiles demonstrated in prior studies (5). We were not able to obtain sufficient detail to categorize bleeding patterns according to the accepted World Health Organization Belsey criteria (18, 19). We also assessed only for the experience of abnormal bleeding at any time during contraceptive implant use and so recall bias may have affected participants’ responses. However, we supported these data with our chart review to obtain objective information regarding treatment for bothersome bleeding. Though 97 participants did not have past medical records available for review, we were able to review records of over 72% of our total participants for this objective analysis. Finally, we only enrolled participants with implants past 12 months of use, and thus did not capture early discontinuers of this method who may have different side effect and bleeding profiles.

The amount of circulating hormone from a contraceptive method can influence the side effects women experience with that method. For the etonogestrel contraceptive implant, high concentrations of the progestin are associated with higher rates of abnormal bleeding and bothersome bleeding requiring medical management. As discontinuation of a contraceptive method places a woman at potential risk for unintended pregnancies, it is imperative to understand all of the clinical factors that contribute to bothersome side effects with hormonal contraception. Our findings highlight that more research is needed on the relationship between pharmacokinetic variability and side effect profiles so that clinicians can continue to improve individualized medication selection and patient counseling on contraceptive options, thereby improving patient satisfaction.