Original research article

Effect of Implanon® use on selected parameters of thyroidand adrenal function

Arijit Biswasa,*, Osborne A.C. Viegasa, Herjan J.T. Coeling Benninkb, Tjeerd Korverb,Shan S. Ratnama

aDepartment of Obstetrics and Gynaecology, National University Hospital, National University of Singapore, 5 Lower Kent Ridge Road, SingaporebNV Organon Research and Development, Oss, The Netherlands

Received 9 November 1999; accepted 12 October 2000

Abstract

In this article, the effects of Implanon® implant use on thyroid and adrenal function was compared with those of Norplant® implants.In this 2-year open randomized study of 80 implant acceptors, we found that both implants may induce minimal changes in thyroid hormonesand cortisol levels, possibly secondary to alterations in the respective binding globulins in the serum. These alterations in thyroid and adrenalfunction would have no clinical significance in healthy women. In the Norplant group, sex hormone-binding globulin levels decreased,whereas increased levels were found in the Implanon users at the end of 2 years. These findings lend support to the fact that etonogestrel,released from Implanon implants, is significantly less androgenic than levonorgestrel, released from the Norplant implants. © 2001 ElsevierScience Inc. All rights reserved.

Keywords:Implanon®; Etonogestrel; 3-Ketodesogestrel; Implants; Progestins

1. Introduction

Implanon® (NV Organon, Oss, The Netherlands) is anew single-rod subdermal contraceptive implant that re-leases etonogestrel, the biologically active metabolite ofdesogestrel. It provides contraceptive protection for 3 yearsand has been approved for use in Indonesia and countries ofthe European Union. Implanon, being a single-rod system,has a simpler insertion and removal procedure. Prior toImplanon, the only other marketed implants were thelevonorgestrel-releasing 6-capsule system, Norplant, andthe two-rod system, Jadelle. The safety and efficacy of thelevonorgestrel-releasing implants are well established.

The objective of the present article was to assess thepossible differences in effects of this new implant system,Implanon, compared to Norplant on selected parameters ofadrenal and thyroid function.

2. Materials and methods

Eighty volunteers were recruited for the study. The vol-unteers were randomized to receive either Implanon orNorplant implants, with 40 subjects in each arm of thestudy. The study period was 24 months. The size of thestudy and control groups are based upon the recommenda-tions made by WHO for metabolic studies, viz., 40 subjectsper group [1].

The women included in the study were between 18–40years of age, sexually active, and of childbearing potential.They had normal menstrual cycles with a mean length ofbetween 24–35 days and an intra-individual variation of63days. They were deemed to have good physical and mentalhealth without contraindications to the use of contraceptivesteroids.

The metabolic parameters were determined on a fastingblood sample, before the insertion of the implant and after6, 12, and 24 months of insertion. The parameters evaluatedwere: total triiodothyronine (T3), total thyroxine (T4), thy-roid binding globulin (TBG), cortisol binding globulin(CBG), testosterone, sex hormone-binding globulin(SHBG) and albumin.

* Corresponding author. Tel.:165-772-4261; fax:165-779-4753.E-mail address:obgab@nus.edu.sg (A. Biswas).

Contraception 62 (2000) 247–251

0010-7824/00/$ – see front matter © 2001 Elsevier Science Inc. All rights reserved.PII: S0010-7824(00)00174-8

T4 and T3 were measured by enzyme immunoassay(EIA) methods using kits from Serono-SRI. TBG levelswere measured by radioimmunoassay (RIA) using kits fromNichols Institute, USA. Plasma cortisol and testosteroneRIA kits were supplied under the auspices of the WHOMatched Reagent Programme. SHBG was measured byimmuno-radiometric assay (IRMA) using kits from Diag-nostic Products Corporation (Los Angeles, CA), whereasCBG was measured by a RIA method using kits fromMedgenix Diagnostics, Fleums, Belgium. Bromocresolgreen (BCG) binding assay method was used to measurealbumin levels, using the COBAS MIRA analyzer and re-agents from Roche Diagnostic Systems, Nutley, New Jer-sey, USA.

Treatment comparison of changes from baseline wereperformed for each assessment and for last measurementsusing the 2-sample Wilcoxon test. Independent samplest-test was used for between-group comparison of means ateach time point.

3. Results

In the Implanon group three subjects (7.5%) and in theNorplant group nine subjects (22.5%) discontinued duringthe study. Summary statistics for absolute values, changesfrom baseline and relative changes from baseline, of thyroidand adrenal function parameters are presented.

3.1. Thyroid function

Levels of thyroid hormones T3 and T4 displayed littlechange from baseline during Implanon use (Table 1). Asimilar result was obtained for Norplant with regard to T3,but T4 was decreased by more than 10% at any treatmentassessment (Fig. 1). Inspection of the absolute mean valuesreveals that this decrease might have resulted from therelatively high baseline T4 values in the Norplant group.Between the two groups, there was no significant differencein the mean T4 levels at any time points.

For TBG, a decrease was observed at 6 months for

Table 1Parameters of thyroid function in both treatment groups

Parameter Implanon® Norplant® Change frombaseline(p-value)a

Mean(p-value)b

N Absolute % Change frombaseline

N Absolute % Change frombaseline

Mean SD Median Mean Median Mean SD Median Mean Median

T3 (nmol/L)Baseline 40 1.60 0.28 1.60 40 1.59 0.37 1.60 0.892Month 6 39 1.45 0.35 1.40 27.62 212.50 37 1.44 0.40 1.40 24.32 26.67 0.938 0.926Month 12 39 1.67 0.37 1.70 6.07 6.25 36 1.65 0.36 1.60 10.89 2.63 0.992 0.845Month 24 37 1.64 0.26 1.60 4.29 0.00 31 1.69 0.31 1.70 10.99 0.00 0.423 0.481Last measurement 39 1.66 0.28 1.70 5.95 0.00 37 1.63 0.34 1.60 6.9925.56 0.876 0.917

T4 (nmol/L)Baseline 40 82.78 17.45 80.0 40 90.08 17.63 94.5 0.066Month 6 39 81.49 16.18 77.0 2.1023.37 37 79.33 15.72 79.0 211.78 213.48 0.009 0.557Month 12 39 82.00 16.28 83.0 4.1022.67 36 80.69 15.53 78.5 210.36 28.82 0.062 0.724Month 24 37 79.19 13.57 79.0 1.7023.75 31 78.84 16.12 79.0 212.37 212.77 0.052 0.923Last measurement 39 79.82 13.49 79.0 1.6723.75 37 78.97 16.45 77.0 212.76 212.77 0.022 0.787

TBG (mg/mL)Baseline 40 25.09 5.07 23.85 40 26.58 4.62 26.58 0.170Month 6 39 23.70 3.19 23.40 24.58 23.68 37 23.58 4.62 22.30 210.40 212.75 0.020 0.901Month 12 39 25.16 3.51 25.80 1.68 1.92 36 24.41 3.52 25.10 26.09 27.42 0.029 0.357Month 24 37 27.32 3.29 26.90 9.75 11.16 31 27.08 5.29 27.10 2.72 1.24 0.034 0.822Last measurement 39 27.31 3.21 26.90 10.46 11.28 37 26.62 5.16 25.60 0.8821.63 0.008 0.540

a Results of the Wilcoxon test on the percentage change from baseline.b Results of independent samples t-test for between-group comparison of means.

Fig. 1. Changes in T4 and T3 levels in implant users (percent changes).

248 A. Biswas et al. / Contraception 62 (2000) 247–251

Implanon compared to baseline, followed by near-baselinevalues at 12 months and increased levels at 24 months. ForNorplant, a decrease was noted at 6 and 12 months, whereasnear-baseline values were reached at 24 months (Table 1).The between-group differences were statistically significantat all assessments.

3.2. Adrenal function

After a decrease from baseline observed at 6 and 12months, median (total) cortisol levels had returned to base-line at 24 months and the last assessment in the Implanongroup (Table 2). In the Norplant group, decreases frombaseline were noted at all time-points (Fig. 2). The between-group differences reached statistical significance at lastmeasurement (p5 0.036). The levels of CBG showedincreases from baseline with both Implanon and Norplant,which were comparable, indicating that the differences inthe cortisol levels cannot be explained by differences in theconcentrations of its major binding protein (Fig. 2).

Implanon and Norplant induced a comparable reductionin (total) testosterone levels. The major binding protein fortestosterone, SHBG, was differently affected by the treat-ments (Fig. 3). With Implanon, SHBG levels were reducedat 6 months and, to a lesser extent, at 12 months, while at 24

months an increase was observed. With Norplant, decreasesfrom the baseline were noted at all assessments, and all werestatistically significantly more prominent than observedwith Implanon.

Albumin levels were slightly increased during both treat-ments, but never more than 5% (Table 3). Group differencesdid not reach statistical significance.

Table 2Parameters of adrenal function in both treatment groups

Parameter Implanon® Norplant® Change frombaseline(p-value)a

Mean(p-value)b

N Absolute Between group N Absolute % Change frombaseline

Mean SD Median Mean Median Mean SD Median Mean Median

Cortisol (nmol/L)Baseline 40 227.53 172.48 225.50 40 238.73 100.16 201.00 0.222Month 6 39 211.41 100.48 191.00 22.60220.39 37 192.54 98.17 166.00 210.91 221.79 0.823 0.411Month 12 39 218.08 99.65 180.00 3.38214.97 36 185.75 91.69 165.00 216.72 230.40 0.282 0.149Month 24 37 253.11 110.28 245.00 20.00 0.87 31 201.58 90.89 174.0028.60 220.30 0.059 0.042Last measurement 39 252.08 108.92 245.00 17.1622.72 37 186.68 90.36 169.00 213.23 224.90 0.036 0.009

Testosterone (nmol/L)Baseline 40 2.19 0.58 2.10 40 2.03 0.63 1.90 0.219Month 6 39 2.02 0.56 1.90 24.84214.29 37 2.07 0.67 1.90 14.77 25.26 0.164 0.705Month 12 39 1.80 0.85 1.60 220.00220.00 36 1.65 0.61 1.60 28.74 217.16 0.445 0.393Month 24 37 1.66 0.58 1.70 223.93223.81 31 1.44 0.47 1.40 220.10 227.27 0.863 0.095Last measurement 39 1.66 0.57 1.70223.64222.58 37 1.47 0.51 1.40 219.67 227.27 0.917 0.443

CBG (nmol/L)Baseline 40 574.68 131.70 625.00 40 570.88 136.21 596.00 0.899Month 6 39 618.59 109.44 608.00 12.54 6.84 37 575.00 79.73 577.00 8.21 2.36 0.547 0.055Month 12 39 645.51 98.87 654.00 17.21 8.62 36 626.08 149.76 620.00 15.15 11.34 0.746 0.716Month 24 37 688.62 71.90 692.00 25.84 11.77 31 656.35 81.03 654.00 21.23 14.31 0.753 0.087Last measurement 39 684.38 73.21 692.00 25.41 11.77 37 671.16 137.33 654.00 25.91 14.84 0.971 0.688

SHBG (nmol/L)Baseline 40 61.53 35.27 53.50 40 53.05 19.94 54.00 0.189Month 6 39 35.36 20.81 29.00 240.48244.12 37 22.57 7.23 22.00 254.37 253.85 0.004 0.001Month 12 39 50.38 32.27 40.00 211.71223.53 36 29.58 9.61 28.00 240.52 246.80 0.004 0.000Month 24 37 72.97 60.59 53.00 23.98 4.49 31 42.13 15.75 37.00213.38 217.07 0.014 0.008Last measurement 39 71.74 59.54 53.00 23.20 4.49 37 40.65 15.21 36.00218.61 225.00 0.002 0.004

a Results of the Wilcoxon test on the percentage change from baseline.b Results of independent samples t-test for between-group comparison of means.

Fig. 2. Changes in plasma cortisol and CBG levels in implant users(percent changes).

249A. Biswas et al. / Contraception 62 (2000) 247–251

4. Discussion

4.1. Effects on thyroid function

The effects of pregnancy and contraceptive steroids, es-pecially estrogen therapy, on laboratory parameters of thy-roid function are well known. The phases of normal men-strual cycle also exert detectable effects [2]. Thyroid studiesperformed during early clinical trials with oral contracep-tives showed the expected changes due to estrogen-inducedrise in TBG. The progestins themselves appeared to have noeffect on these laboratory parameters [3,4]. Progestin-onlypills do not alter conventional plasma thyroid tests [5].Croxatto et al. [6] studied levels of T4, T3, and TSH in 2groups of women, one using levonorgestrel-releasing Nor-plant implants, and the other using a copper intrauterinedevice (IUD). They reported slightly lower levels of T3 inthe Norplant implant group than in the IUD group. T4 andTSH levels did not differ significantly between the 2 groups.However, this was a cross-sectional study with samplingdone after 37 months of use. In a longitudinal study over 6

months, Olsson et al. [4] found no change in thyroid func-tion, except for the impact of a slight decrease in TBG level.

In the present comparative study with Implanon andNorplant implants, no changes were noted in the levels ofthe thyroid hormones T3 and T4 in the Implanon group. Inthe Norplant group, there was no change in T3 levels, but T4was slightly decreased during treatment. However, therewas no significant difference seen on between-group com-parison of mean T4 at any time point. TBG levels weretransiently decreased with both treatments, but less pro-nounced in the Implanon group, although there was nosignificant difference in the mean TBG values at any time.This marginal difference in TBG levels may be secondary tothe lower intrinsic androgenicity of etonogestrel in Impla-non compared with that of levonorgestrel in Norplant im-plants. It is important to note that at no time points were thethyroid function parameters in either group, outside thenormal range in our population.

4.2. Effects on adrenal function

Synthetic progestins exhibit glucocorticoid-like activitythat varies with the type of progestational agent and itsconcentration. The glucocorticoid activity may be of suffi-cient magnitude to exert a negative feedback effect on theanterior pituitary, and cause a decrease in ACTH levels aswell as lower cortisol levels. Natural progesterone is appar-ently devoid of glucocorticoid and negative feedback effect.In the present comparative study, mean cortisol levelstended to increase in Implanon users and were decreased inNorplant users. However, between-group difference wasstatistically significant only at the last (24 months) measure-ment.

The results of the present study suggest that etonogestrel,released from the Implanon implants, has negligible glu-cocorticoid activity and a negligible negative feedback ef-fect on the pituitary. The increase in cortisol levels seen inImplanon users could be explained by a concomitant rise inCBG. CBG levels were increased in both groups. Ninetypercent of plasma cortisol is bound to CBG. This protein hashigh affinity but limited binding capacity. Complete satura-

Fig. 3. Changes in serum testosterone and plasma SHBG levels in implantusers (percent changes).

Table 3Serum albumin concentrations in both treatment groups

Parameter N Implanon® N Norplant® p-valuea

Absolute % Change frombaseline

Absolute % Change frombaseline

Mean SD Mean Median Mean SD Mean Median

Albumin (g/L)Baseline 40 45.70 3.11 40 44.78 2.19Month 6 39 46.28 3.91 1.36 2.04 37 46.03 3.03 3.06 2.33 0.560Month 12 39 46.82 3.05 2.48 2.27 36 46.31 2.30 3.45 4.35 0.671Month 24 37 46.27 3.40 1.14 2.22 31 47.06 2.61 4.64 4.35 0.083Last measurement 39 46.41 3.40 1.55 2.22 37 46.89 2.63 4.68 4.35 0.063

a Results of the Wilcoxon test on the percent change from baseline.

250 A. Biswas et al. / Contraception 62 (2000) 247–251

tion occurs at a plasma level of 20mg/dL with about 10%free cortisol. At higher concentrations, free cortisol in-creases nonlinearly to about 30%, with the remainder beingbound to albumin, which has low affinity but high capacity[7]. Our results showed that albumin levels were minimallybut non-significantly increased in both treatment groups(Table 3).

Only 1% to 3% of the plasma testosterone circulates inthe biologically active, unbound form. Two circulating mac-romolecules bind sex steroids: 1) albumin, a low-affinity,high-capacity, binding protein and 2) sex hormone-bindingglobulin (SHBG), a high-affinity, low-capacity binding pro-tein that binds testosterone preferentially over estradiol [7].In women, SHBG levels are 2-fold greater than in men, andtestosterone concentrations are much lower. Thus, inwomen, SHBG has many available binding sites. Estrogensstimulate and androgens inhibit SHBG production. Differ-ent progestational agents, depending on their intrinsic an-drogenicity, have different effects on SHBG production.Progesterone and synthetic pregnane derivatives do nothave an effect on SHBG production, but synthetic proge-stagens derived from 19-nortestosterone, decrease plasmaSHBG levels. In the present study, compared with the Im-planon group, SHBG levels were statistically significantlymore decreased in the Norplant group, at all assessments.While in the Norplant group decreases from the baselinewere seen at all time-points, in the Implanon users near-baseline values were seen at 12 months and increased levelsat 24 months. These findings are in accordance with the factthat etonogestrel, released from Implanon implants, is sig-nificantly less androgenic than levonorgestrel, released fromthe Norplant implants. Total testosterone levels were re-duced in both groups. This is believed to be mainly due toa fall in ovarian steroid production.

In summary, it appears that both implants have no sig-nificant effect on thyroid function and may induce onlyminimal changes in cortisol levels, possibly secondary toalterations in the binding globulins in the serum. Theseslight alterations in adrenal function would have no clinicalsignificance in the healthy woman.

Acknowledgements

The authors would like to thank Ms Jamila Beevi, thestudy co-ordinator, and the staff of the Fertility ControlCenter, National University Hospital, Singapore, for theirsupport and help.

References

[1] Michal F. Safety Requirements for Contraceptive Steroids. Publishedon behalf of WHO. UK: Cambridge University Press, 1989:439.

[2] Beck RP, Fawcett DM, Morcos F. Thyroid function studies in dif-ferent phases of the menstrual cycle and in women receiving noreth-indrone with and without estrogen. Am J Obstet Gynecol 1972;112:369–73.

[3] Winikoff D. Oral contraceptives and thyroid function tests: the role ofprogestogens. Med J Aust 1968;2:13–8.

[4] Olsson SE, Wilde L, Odlind V. Aspects of thyroid function during theuse of Norplant implants. Contraception 1986;34:583–7.

[5] Goolden AW, Bateman DM, Pleehachinda R, Sanderson C. Thyroid-function tests in women taking norgestrel. Lancet 1970;1:624.

[6] Croxatto HB, Diaz S, Robertson DN, Pavez M. Clinical chemistry inwomen treated with levonorgestrel implants (Norplant) or a TCu 200IUD. Contraception 1983;27:281–8.

[7] Anderson DC. Sex-hormone binding globulin. Clin Endocrinol 1974;3:69–96.

251A. Biswas et al. / Contraception 62 (2000) 247–251