- Email: [email protected]
Ovarian endocrine function through five years of continuous treatment with NORPLANT® subdermal contraceptive implants
CONTRACEPTION
OVARIAN ENDXRINE FUNCTIaNTHROUGH FIVE YEARS OF CONTINUOUSTREATMENT WITH NORPLANT SUBDERMALCONTRACEPTIVEIMPLANTS Brache V*, Alvarez-SanchezF*, FaundesA**, Tejada AS*, Cochon L*
*CINSERHA- PROFAMILIA Socorro Sanchez 64 - Santo Domingo,DominicanRepublic **Departmentof Obstetricsand Gynecology,Faculty of Medical Sciences, UniversidadeEstadualde Campinas, PopulationCouncil Office, Caixa Postal 6181 - CEP 13081 - Campinas,SP, Brazil
Ovarian endocrine function was assessed in 88 women using NORPLANTR subdermal implants during different periods of use and in a control group of 15 women using non-hormonalcontraception. Blood samples for estradiol (E2) and progesterone (P) assays were obtained twice a week for five consecutive weeks. Three distinct E patterns were observed: one was characterizedby fluctuating levels wit&n a normal range (20 to 400 &ml), a second pattern corresponded to continuous low E levels (below 75 pg/ml in the 10 samples) and the third was characterizedb y high broad estradiol peaks reaching over 400 pg/ml. The proportion of sampling runs characterized by normal fluctuating levels increased from 38% in the first two years of use to 80% during the fifth year of use. Low E profile was only observed during the first two years of use (27%) and in on1y 1 case at the beginningof the third year of use (5%). The percentage of cycles with high broad estradiol peaks remained between 20-40% without a clear tendencyto change in either direction with duration of use. Thirty-threepercent of the observed samplingruns had luteal activity (P above 3 ng/ml). The proportionof runs with luteal activity increasedfrom 14% during the first two years of use to 40% during the third and fourth, and 60% during the fifth year of use. All control subjectshad luteal activity. The mean highest progesteronelevel was lower in the NoRPLANTRruns (8.7 +_3.9 ng/ml) as compared to the controls (11.3 + 3.8 ng/ml). NORPLANTR sampling runs with luteal activity had normal fluctuatingE2 levels with only one exception. However,not all cycles with normal E2 levels showed luteal activity. On the other hand, all runs with low E2 levels or high broad E2 peaks were without luteal activity. In summary, women using continuous low-doselevonorgestrelcontraception through NCRPLANTR subdermal implants, have a variable degree of ovarian activity as compared with the more complete depression of ovarian function observed among pill or injectables'users. Ovarian activity becomes closer to normal during the third through fifth year of use. Submittedfor publicationJuly 12, 1989 Accepted
for publication
October
FEBRUARY 1990 VOL. 41 NO. 2
27, 1989
169
CONTRACEPTION
The concept of hormonal contraceptionwith the continuousadministration of low doses of progestagenshas been accepted and practicedduring the last 20 years, mostly in the form of the so-calledminipill. The mOre effective application of the same concept by the administrationof progestagens in subdermal implants is more recent, but has already reached over 200,000 users and several thousandwoman/yearsof clinicalobs rvation in Phase III or field studies of the first commercialproduct NORPLAt?#*(l). Although several authorsPhave reported data on ovarian and pituitary function during use of NORPLANT , most of the data have been restrictedto the first year of use (2,3,4). Croxatto et al. howeve did study progesteroneand estradiol levels after long-term use of NORPLANIJ(5r6). Our group has also reported luteal functionamong long-termusers of NORPLANTR,but in a selected sub-group of women with normal cyclic menstrualpattern. This la ter group only comprises about half of the subjectsduring prolongedMXPL ANI! use (7). We have also previously detailed the FSH, LH, estradiol and progesterone profile in an indeptsstudy of a small group of women with luteal activity during use of NORPLANT (8). The aim of the present study is to provide a general pi ure of the ovarian endocrine function in a sample of wcmen using NORPLAN% with any menstrual pattern and in various periods during the five years of use.
Estradiol (E ) and progest one (P) levels were measured in 10 serum users during five consecutive weeks (2 samples collected2 from NCRPL3 samples per week). Women participatingin Phase III NORPLANTRclinical trials conducted at the Centro de Investigation y Servicios en ReproducciGn Humana Y AnticoncepciBn (CINSERHA/PROFAMILIA)in Santo Domingo, Dominican Republic, were informed of this protocol at their regular follow-upvisits, and those who volunteered to participate initiated sampling at the same visit, regardlessof the day of the cycle or of their bleedingpattern.
* NORPLANTR is the registeredtrademarkof the PopulationCouncil for subdermalcontraceptiveimplants.
170
FEBRUARY 1990 VOL. 41 NO. 2
CONTRACEPTION
A control group of 15 subjects using non-hormonalcontraceptiveswere enrolled followingthe same procedure. R. Eighty-eightsubjects using WORPLAWT implants participated,yielding a total of 92 sampling runs: 17 in the first year of use, 20 in the second, 20 in the third, 20 in the fourth and 15 in the fifth year of use. Four subjects participatedin two sampling runs, but in differentperiods of use. Blood samples were allowed to clot, and centrifugedwithin 2 hrs. The serum was stored at -20 C until analysis. To minimize interassayvariation, all samples from a single subject (ie. samplingrun) were analyzed in the same radioimnunoassay. 125 Estradiol and progesterone assays were performed by a solid-phaseI technique with a sensitivity of 10 pg/ml and 0.05 ng/ml, and an interassay coefficientof variationof 10.0 and 12.1%, respectively. Each assay included both commercial quality control sera, and control sera of the WHO Matched Reagents Progrannne for the Radioimmunoassay of Hormones in Human Reproduction. The data were analyzed using each 5-week sampling run as the unit for analysis. Each sampling run was classified according to the dominant characteristicof the E and P levels. Wheneverserum P was above 3 ng/ml in one or more samples, thaz run was classifiedas having luteal activity. RESULTS
Estradiol: There were three distinctpatterns of serum estradiollevels. One was characterizedby cyclic fluctuationswithin a normal range (up to 400 pg/ml): a second pattern correspondedto continuouslylow levels (below 75 p&l in the 10 samples);andthe third was characterizedby fluctuatinglevels but with high broad peaks, above 400 pg/ml, sometimes sustained for up to 10 days (three consecutivesamples). TabAe I shows the distributionof the three patternsalong the 5 years of NORPLANT usage. With the exceptionof the second year of use, the estradiol pattern most frequentlyobserved was fluctuatingE levels within the normal was significantly range. The proportion of cycles with normal lower during ty first two years of use (38%), o 71% during years 3 through 5 (X 9.90, peO.005). The proportionof cycles showing E2 peaks above 400 pg/ml remainedbetween 20 and 40% without a clear tendencyto change in either direction with duration of use. Low E2 profiles were only observed during the first two years and early in the third year of use (Month 26). Table II shows the individualE2 values for these subjects. All control subjects had E2 levels within the normal range with the exceptionof one case that had a pre-ovulatorypeak above 400 pg/ml.
FEBRUARY X390 VOL. 41 NO. 2
171
CONTRACEPTION
TABLE I DISTRIBUTIONOF ESTRADIOLPATTERNSAMONG NoRPLANTRJSERS AND CONTROLS
SUSTAINEDLEVELS BELOW 75 &ml N % 3/17 7/20 l/20 o/20 o/15
FIRSTYEAR SECONDYEAR THIRD YEAR FOURTH YEAR FIFTH YEAR
u/92
!lWl'AL
BROAD PEAKS ABOVE 400 w/ml N %
18 35 5 0 0
8/17 6/20 15/20 12/20 12/15
47 30 75 60 80
6/17 7/20 4/20 8/20 3/15
35 35 20 40 20
12
53/92
58
28/92
30
0
14/15
93
l/15
7
o/15
CONTROLS
NORMALRANGE (20400 pg/ml) % N
TABLE II INDIVIDUALES'ITUDIOL LEVELS (PG/ML)IN SUBJECTSWITH SUSTAINEDLOW E
No.
MONTH OF USE
T-459 T-486 ~-323 T-323 346 T-118 342 PIT-325 ~-216 T-230 T-126
172
2 2 8 13 13 13 14 14 14 14 26
LEVELS (< 75 PG/ML) DURItbG NORPLANTRUSE
SAMPLES (5 CONSECUTIVEWEEKS/2 SAMPLES PER WEEK), 2 3 4 5 6 7 8 9 10 MEAN 38 34 26 18 73 28 10 21 57 73 34
43 31 36 42 59 31 19 24 59 63 36
54 43 35 50 10 41 21 36 45 54 33
42 31 16 43 35 21 10 30 68 37 38
41 39 30 23 25 30 38 28 41 35 32
63 41 38 23 11 52 10 36 56 39 32
62 48 20 27 26 43 28 10 49 44 34
35 42 22 31 10 22 20 10 50 34 28
50 46 22 37 23 28 16 15 45 65 35
64 42 15 41 54 10 16 64 48 49
49.2 39.7 26.0 33.5 32.6 32.9 18.2 22.6 53.4 49.2 32.7
FEBRUARY 1990VOL. 41 NO. 2
CONTRACEPTION
Progesterone: Plasma progesteroneabove 3 ng/ml was estimated to be an indication of luteal activity. Two-thirds of the cycles did not show any luteal activity. The proportionwith P above 3 ng/ml. increased from one-fifth (18%) and one-tenth of the cycles in the first two years of use, to more than half of the cycles in the fifth year (60%) (~~0.025). The difference in luteal activity by length of use is more evident if the proportion with P above 6 ng/ml is analyzed,as shown in Table III. All control subjects had luteal activity expressedby P levels above 3 ng/ml a in all but one, P levels were above 6.ng/ml. P levels during use of NORPLAN? were significantlylower, as compared with the control group. The mean highest progesteronelevel in the NORPLANI?group was 8.7 f 3.9 vs. 11.3 + 3.8 in the controls (p 40.025). This has been described previously (4,5,7,8). TABLE III PROPORTIONOF CYCLES WITH LUTEAL ACTIVITYDURING USE OF NORPLANTR
YSAR OF USE
PROPORTIONWITH PRCGESTRRONELEVELS ABOVE 3 ng/ml ABOVE 6 ng/ml 3/17 2/20 8/20 8/20 9/15
18% 10% 40% 40% 60%
l/17 l/20 5/20 6/20 8/15
6% 5% 25% 30% 53%
XCTAL
30/92
33%
21/92
23%
CONTROLS
15/15
100%
14/15
93%
1st 2nd 3rd 4th 5th
Luteal Activity and Estradiol Pattern: Cycles showing luteal activity corresponded to thosewith E2 levmhin the normal range. The one exceptionhad an E2 pre-ovulatorypeak of 417 pg/ml and mid-lutealpeak of 410 &ml. However, not all the cycles with E2 levels within the normal range showed luteal activity. The proportionof cycles with normal E2 and luteal activity increasedwith the time of use, from 25% in the first year to 75% in the 5th year of use (Table IV). The evolut' n of the ovarian endocrine function along the five years of use of NORPLAN is sursnarizedin Fig. 1. The proportion of cycles with normal E2 and luteal activity increasedwith time. Conversely,cycles without luteal activity and sustained low E2 levels decreased. There was less change in the percentage of cycles without luteal activity and E2 either within normal range or with peaks above 400 pg/ml.
FEBRUARY
1990 VOL. 41 NO. 2
173
CONTRACEPTION
TABLE IV PROPORTIONOF CYCLES WITH LUTEAL ACTIVITYAMONG NORPLANTRUSEPuS WITH MlRMAL !3STRADIOL YEARS OF USE
LUTEAL ACTIVITY
1st 2nd 3rd 4th 5th
2/8 2/6 8/15 8/12 9/12
25% 33% 53% 67% 75%
CONTROLS
14/14
100%
E2 NORMAL E2 NORMAL / P ) 14%
/ P ) 3
2 LOW I P ( 3 27% E2 LOW I P ( 3 2%
E2 NORMAL 25%
/ P <3 E2 ) 400 / P ( 3 27%
~4OO,P~3 32%
YEAR
1 AND
2
YEAR
3, 4, AND 5
Figure 1: Ovarian endocrine rofiles observedduring differentperiods of use of NORPLANTa subdernal implants.
174
FEBRUARY 1990 VOL. 41 NO. 2
CONTRACEPTION
DISUJSSIDN The most obvious advantage of continuous low-dose progestagen contraceptionover other hormonal methods is the smaller amount of exogenous hormones received by the users. The expectation is to obtain the same efficacy without the side effects which occur with higher doses. In fact, the history of hormonal contraceptionhas been characterizedby the progressive The low doses of reduction in the amount of administered hormones. progestagensand estrogensin the more recent pill formulations,have resulted in fewer side effects and lower risks of severe complications. A method with even lower daily dose and circulating levels of progestagens,without any estrogen,would be expectedto have even less health risks for the users. The effect of the exogenous hormones on circulatingsteroids is at least part of the intended contraceptiveeffect. The main differencebetween the continuous low-dose progestagen-onlysubdermal implantsand the combinedpill or depot-injectables,is that the first is not necessarilyanovulatory,but still is as effectiveor even more effectivethan the latter. Sev ral studies have shown endocrinological changes during use of NORPLAN?!. High and broad E peaks, anovulationand luteal insufficie y have all been describedas part oz the ovarian endocrinepicture in NORPLAN? users (2,3,4,5,6,7,8).What was still lacking was a more precisedescriptionof the quality, frequencyand dimensionof the changes and their variationalong time of use. The present study shows that, as expected,not all NORPLAFI?users have the same ovarian endocrine response to the continuous exposure to levonorgestrel. The observed effect varies from an apparently profound depressionof ovarian function,with persistentlylow E2 levels without luteal activity in some cycles, in some women, to practicallyno change as compared to normal cyclic E2 and P in other women. In one-thirdof the cycles studied, luteal activitywas observed. Length of use has a clear influenceon luteal function. Not only the proportionof subjects with luteal activityincreaseswith time but higher P levels are more frequentlyobserved in the fourth and fifth year of use. On the other hand, low E2 levels foE periods of 30 or more days, had already been found among users of NORPLANT (3,6). Accordingto our results, this happens in about one out of every four cycles during the first two years of use, in one out of 20 in the third year of use, and is not observed thereafter. However, the method used in this study, e.g. 10 samples in a 5-week period, is acceptable to determine the proportion of cycles with luteal activity but not necessarilyto define persistentlylow E2 secretion. We observed only a relativelysmall "window" of the general ovarian function:a different view could be obtainedby prolongingthe sampling for more weeks. Our results illustratethe effect of length of samplingover the apparent findings: in addition to the 11 subjects with E2 below 75 pg/ml in all samples, we had 12 with the same low levels in 8 or 9 of the 10 samples, meaning that if we had limited the sampling to 4 weeks, we would have artificiallydoubled the proportionwith low E2 levels.
FEBRUARY 1990 VOL. 41 NO. 2
175
CONTRACEPTION
Similarly, Croxatto et al. (6) found only one out of 61 cycles studied with all E2 levels below 100 pg/ml., but some of their samplingruns covered six weeks of observation instead of five. Their study included a smaller number of cycles within the first years of use (11/61 as comparedto 37/92 in our study), which may also contribute to explain the difference with our results. Presently,persistentlylow E2 levels (below40 &ml) are automatically linked to a higher risk of osteoporosis. Only three of our 92 observed cycles maintained levels below 40 pg/ml during the sampling period. The weight of these three patients was very low (41, 43 and 54 Kg), thereforethese women probably had greater ovarian su@oressiondue to higher levonorgestrellevels, as has been observed in NORPLANT users with low body weight (1). The question as to how long these low E2 levels are in fact maintained, would be answered only with a much more prolonged study, with all the practical problems it would present. However, the real dimensionof such risk may be reduced or eliminatedin these women since they are under continuous progestin administration,which has ,been shown to have a protectiveeffect against bone loss (9, 10). The presence of occasionalE peaks, prolonged for up to 10 days, could also seem alarming, considering w e absence of luteal activity in all those cycles. It does not, however,constitutea situationof unopposedestrogenic activity as observed in other clinical confflitions, as for example polycystic ovary syndrome. In the case of NORPLANT users, the subject's tissues are simultaneously under the continuous effect of a progestagen with potent antiestrogeniceffect. Moreover, we did not find any period longer than 10 days (three samples)with E2 over 400 pg/ml. In continuous low-dose, progestagen-only summary, women u@ng contraceptionthrough NORPLANT , have a variabledegree of endocrineactivity as compared with the complete depression of ovarian function among pill or injectables'users. Ovarian activity&ecomes closer to normal during the 3rd through 5th year of use of NORPLANT . This lesser depression of ovarian endocrine function explains the rapid resumption of ovulation (11) and recovery of fertilityafter discontinuationof use by removalof the implants (12, 13). Acknowledgements: This study was undertaken as part of the contraceptivedevelopment program sponsored and coordinated by the International Committee for ContraceptionResearch of the Population Council, New York. The financial support provided by the US Agency for International Development, the RockefellerFoundation,the George Hecht Fund, the Andrew W. Mellon Foundation and the United Nations Fund for Population Activities is gratefully acknowledged.
176
FEBRUARY 1990 VOL. 41 NO. 2
CONTRACEPTION
REFERENCES 1.
Sivin I. International Experience with NORPLANTR and NORPLANTContraceptives. Studies in Family Planning19:81-94 (1988).
2.
Weiner E, Johansson EDB. Plasma Levels of d-Norgestrel,Estradiol and Progesterone during Treatment with Silastic Implants Containing d-Norgestrel. Contraception14:81-92 (1976).
3.
Bleeding and Serum Moore DE, Roy s, Stanczyk FZ, Mishell DR. d-Norgestrel, Estradiol and Progesterone Patterns in Women Using d-Norgestrel Subdennal Polysiloxane Capsules for Contraception. Contraception17:315-328(1978).
4.
Shaaban MM, El-NasharIM, GhaneimahSA, Gomaa AA, Salah M, Abdel-AleemA. Hormonal Changes ring the First Year of Use of SubdermalLevonorgestrel Implants,NORPLAN? . Contraception30:391-405(1984).
5.
Croxatto HB, Diaz S, Paves M, Miranda P, BrandeisA. Plasma Progesterone Levels During Long-term Treatment with LevonorgestrelSilastic Implants. Acta Endocrinol.101:307-311(1982).
6.
Croxatto HB, Diaz S, Pavez M, Brand 's A. EstradiolPlasma Levels During Long-Term Treatment with NDRPLAN? Subdermal Implants. Contraception 38:465-475(1988).
7.
Brache V, Faundes A, Johansson E, Alvarez F. Anovulation, Inadequate Luteal Phase and Poor Sperm Penetration in Cervical Mucus During ProlongedUse of NORPLANTRImplants. Contraception31:261-273(1985).
8.
Alvarez F, &ache V, Tejada AS, Faundes A. Abnormal Endocrine Profile Among Wmen with Confirmed or Presumed Ovulation During Long-Term NORPLANTa Use. Contraception33:111-119(1986).
9.
Lindsay R, Hart DM, Purdie D et al. ComparativeEffects of Estrogen and Progesterone on Bone Loss in PostmenopausalWomen. Clin Sci Mol Med 54:193 (1978).
Effect of 10. Davidson BJ, Deftos LJ, Meldrum DR, Judd HL. MedroxyprogesteroneAcetate (MPA) on Bone Metabolism of Postmenopausal Women. Abstr No.279, Society for GynecolcgicInvestigation,27th Annual Meeting, Denver,CO, March 1980. Ovulation Detection 11. Ismail AAA, Anwar MY, Youssef SM, Toppozada M. Following Removal of LevonorgestrelSubdermal Contraceptive Implants. Contraception35:207-214(1987). 12. Affandi B, santoso SSI, Dapjadilaga HadisaputraW et al. Pregnancy After Removal of NORPLANT Implants Contraceptive. Contraception 36:203-209(1987). 13. Diaz S, Paves M, Cardenas H, Croxatto HB. Recovery of Fertility and Outcome of Planned PregnanciesAfter the Removal of NORPLANTR Subdermal Implants or Copper-T IUDs. Contraception35:569-579(1987).
FEBRUARY 1990 VOL. 41 NO. 2
177
OVARIAN ENDXRINE FUNCTIaNTHROUGH FIVE YEARS OF CONTINUOUSTREATMENT WITH NORPLANT SUBDERMALCONTRACEPTIVEIMPLANTS Brache V*, Alvarez-SanchezF*, FaundesA**, Tejada AS*, Cochon L*
*CINSERHA- PROFAMILIA Socorro Sanchez 64 - Santo Domingo,DominicanRepublic **Departmentof Obstetricsand Gynecology,Faculty of Medical Sciences, UniversidadeEstadualde Campinas, PopulationCouncil Office, Caixa Postal 6181 - CEP 13081 - Campinas,SP, Brazil
Ovarian endocrine function was assessed in 88 women using NORPLANTR subdermal implants during different periods of use and in a control group of 15 women using non-hormonalcontraception. Blood samples for estradiol (E2) and progesterone (P) assays were obtained twice a week for five consecutive weeks. Three distinct E patterns were observed: one was characterizedby fluctuating levels wit&n a normal range (20 to 400 &ml), a second pattern corresponded to continuous low E levels (below 75 pg/ml in the 10 samples) and the third was characterizedb y high broad estradiol peaks reaching over 400 pg/ml. The proportion of sampling runs characterized by normal fluctuating levels increased from 38% in the first two years of use to 80% during the fifth year of use. Low E profile was only observed during the first two years of use (27%) and in on1y 1 case at the beginningof the third year of use (5%). The percentage of cycles with high broad estradiol peaks remained between 20-40% without a clear tendencyto change in either direction with duration of use. Thirty-threepercent of the observed samplingruns had luteal activity (P above 3 ng/ml). The proportionof runs with luteal activity increasedfrom 14% during the first two years of use to 40% during the third and fourth, and 60% during the fifth year of use. All control subjectshad luteal activity. The mean highest progesteronelevel was lower in the NoRPLANTRruns (8.7 +_3.9 ng/ml) as compared to the controls (11.3 + 3.8 ng/ml). NORPLANTR sampling runs with luteal activity had normal fluctuatingE2 levels with only one exception. However,not all cycles with normal E2 levels showed luteal activity. On the other hand, all runs with low E2 levels or high broad E2 peaks were without luteal activity. In summary, women using continuous low-doselevonorgestrelcontraception through NCRPLANTR subdermal implants, have a variable degree of ovarian activity as compared with the more complete depression of ovarian function observed among pill or injectables'users. Ovarian activity becomes closer to normal during the third through fifth year of use. Submittedfor publicationJuly 12, 1989 Accepted
for publication
October
FEBRUARY 1990 VOL. 41 NO. 2
27, 1989
169
CONTRACEPTION
The concept of hormonal contraceptionwith the continuousadministration of low doses of progestagenshas been accepted and practicedduring the last 20 years, mostly in the form of the so-calledminipill. The mOre effective application of the same concept by the administrationof progestagens in subdermal implants is more recent, but has already reached over 200,000 users and several thousandwoman/yearsof clinicalobs rvation in Phase III or field studies of the first commercialproduct NORPLAt?#*(l). Although several authorsPhave reported data on ovarian and pituitary function during use of NORPLANT , most of the data have been restrictedto the first year of use (2,3,4). Croxatto et al. howeve did study progesteroneand estradiol levels after long-term use of NORPLANIJ(5r6). Our group has also reported luteal functionamong long-termusers of NORPLANTR,but in a selected sub-group of women with normal cyclic menstrualpattern. This la ter group only comprises about half of the subjectsduring prolongedMXPL ANI! use (7). We have also previously detailed the FSH, LH, estradiol and progesterone profile in an indeptsstudy of a small group of women with luteal activity during use of NORPLANT (8). The aim of the present study is to provide a general pi ure of the ovarian endocrine function in a sample of wcmen using NORPLAN% with any menstrual pattern and in various periods during the five years of use.
Estradiol (E ) and progest one (P) levels were measured in 10 serum users during five consecutive weeks (2 samples collected2 from NCRPL3 samples per week). Women participatingin Phase III NORPLANTRclinical trials conducted at the Centro de Investigation y Servicios en ReproducciGn Humana Y AnticoncepciBn (CINSERHA/PROFAMILIA)in Santo Domingo, Dominican Republic, were informed of this protocol at their regular follow-upvisits, and those who volunteered to participate initiated sampling at the same visit, regardlessof the day of the cycle or of their bleedingpattern.
* NORPLANTR is the registeredtrademarkof the PopulationCouncil for subdermalcontraceptiveimplants.
170
FEBRUARY 1990 VOL. 41 NO. 2
CONTRACEPTION
A control group of 15 subjects using non-hormonalcontraceptiveswere enrolled followingthe same procedure. R. Eighty-eightsubjects using WORPLAWT implants participated,yielding a total of 92 sampling runs: 17 in the first year of use, 20 in the second, 20 in the third, 20 in the fourth and 15 in the fifth year of use. Four subjects participatedin two sampling runs, but in differentperiods of use. Blood samples were allowed to clot, and centrifugedwithin 2 hrs. The serum was stored at -20 C until analysis. To minimize interassayvariation, all samples from a single subject (ie. samplingrun) were analyzed in the same radioimnunoassay. 125 Estradiol and progesterone assays were performed by a solid-phaseI technique with a sensitivity of 10 pg/ml and 0.05 ng/ml, and an interassay coefficientof variationof 10.0 and 12.1%, respectively. Each assay included both commercial quality control sera, and control sera of the WHO Matched Reagents Progrannne for the Radioimmunoassay of Hormones in Human Reproduction. The data were analyzed using each 5-week sampling run as the unit for analysis. Each sampling run was classified according to the dominant characteristicof the E and P levels. Wheneverserum P was above 3 ng/ml in one or more samples, thaz run was classifiedas having luteal activity. RESULTS
Estradiol: There were three distinctpatterns of serum estradiollevels. One was characterizedby cyclic fluctuationswithin a normal range (up to 400 pg/ml): a second pattern correspondedto continuouslylow levels (below 75 p&l in the 10 samples);andthe third was characterizedby fluctuatinglevels but with high broad peaks, above 400 pg/ml, sometimes sustained for up to 10 days (three consecutivesamples). TabAe I shows the distributionof the three patternsalong the 5 years of NORPLANT usage. With the exceptionof the second year of use, the estradiol pattern most frequentlyobserved was fluctuatingE levels within the normal was significantly range. The proportion of cycles with normal lower during ty first two years of use (38%), o 71% during years 3 through 5 (X 9.90, peO.005). The proportionof cycles showing E2 peaks above 400 pg/ml remainedbetween 20 and 40% without a clear tendencyto change in either direction with duration of use. Low E2 profiles were only observed during the first two years and early in the third year of use (Month 26). Table II shows the individualE2 values for these subjects. All control subjects had E2 levels within the normal range with the exceptionof one case that had a pre-ovulatorypeak above 400 pg/ml.
FEBRUARY X390 VOL. 41 NO. 2
171
CONTRACEPTION
TABLE I DISTRIBUTIONOF ESTRADIOLPATTERNSAMONG NoRPLANTRJSERS AND CONTROLS
SUSTAINEDLEVELS BELOW 75 &ml N % 3/17 7/20 l/20 o/20 o/15
FIRSTYEAR SECONDYEAR THIRD YEAR FOURTH YEAR FIFTH YEAR
u/92
!lWl'AL
BROAD PEAKS ABOVE 400 w/ml N %
18 35 5 0 0
8/17 6/20 15/20 12/20 12/15
47 30 75 60 80
6/17 7/20 4/20 8/20 3/15
35 35 20 40 20
12
53/92
58
28/92
30
0
14/15
93
l/15
7
o/15
CONTROLS
NORMALRANGE (20400 pg/ml) % N
TABLE II INDIVIDUALES'ITUDIOL LEVELS (PG/ML)IN SUBJECTSWITH SUSTAINEDLOW E
No.
MONTH OF USE
T-459 T-486 ~-323 T-323 346 T-118 342 PIT-325 ~-216 T-230 T-126
172
2 2 8 13 13 13 14 14 14 14 26
LEVELS (< 75 PG/ML) DURItbG NORPLANTRUSE
SAMPLES (5 CONSECUTIVEWEEKS/2 SAMPLES PER WEEK), 2 3 4 5 6 7 8 9 10 MEAN 38 34 26 18 73 28 10 21 57 73 34
43 31 36 42 59 31 19 24 59 63 36
54 43 35 50 10 41 21 36 45 54 33
42 31 16 43 35 21 10 30 68 37 38
41 39 30 23 25 30 38 28 41 35 32
63 41 38 23 11 52 10 36 56 39 32
62 48 20 27 26 43 28 10 49 44 34
35 42 22 31 10 22 20 10 50 34 28
50 46 22 37 23 28 16 15 45 65 35
64 42 15 41 54 10 16 64 48 49
49.2 39.7 26.0 33.5 32.6 32.9 18.2 22.6 53.4 49.2 32.7
FEBRUARY 1990VOL. 41 NO. 2
CONTRACEPTION
Progesterone: Plasma progesteroneabove 3 ng/ml was estimated to be an indication of luteal activity. Two-thirds of the cycles did not show any luteal activity. The proportionwith P above 3 ng/ml. increased from one-fifth (18%) and one-tenth of the cycles in the first two years of use, to more than half of the cycles in the fifth year (60%) (~~0.025). The difference in luteal activity by length of use is more evident if the proportion with P above 6 ng/ml is analyzed,as shown in Table III. All control subjects had luteal activity expressedby P levels above 3 ng/ml a in all but one, P levels were above 6.ng/ml. P levels during use of NORPLAN? were significantlylower, as compared with the control group. The mean highest progesteronelevel in the NORPLANI?group was 8.7 f 3.9 vs. 11.3 + 3.8 in the controls (p 40.025). This has been described previously (4,5,7,8). TABLE III PROPORTIONOF CYCLES WITH LUTEAL ACTIVITYDURING USE OF NORPLANTR
YSAR OF USE
PROPORTIONWITH PRCGESTRRONELEVELS ABOVE 3 ng/ml ABOVE 6 ng/ml 3/17 2/20 8/20 8/20 9/15
18% 10% 40% 40% 60%
l/17 l/20 5/20 6/20 8/15
6% 5% 25% 30% 53%
XCTAL
30/92
33%
21/92
23%
CONTROLS
15/15
100%
14/15
93%
1st 2nd 3rd 4th 5th
Luteal Activity and Estradiol Pattern: Cycles showing luteal activity corresponded to thosewith E2 levmhin the normal range. The one exceptionhad an E2 pre-ovulatorypeak of 417 pg/ml and mid-lutealpeak of 410 &ml. However, not all the cycles with E2 levels within the normal range showed luteal activity. The proportionof cycles with normal E2 and luteal activity increasedwith the time of use, from 25% in the first year to 75% in the 5th year of use (Table IV). The evolut' n of the ovarian endocrine function along the five years of use of NORPLAN is sursnarizedin Fig. 1. The proportion of cycles with normal E2 and luteal activity increasedwith time. Conversely,cycles without luteal activity and sustained low E2 levels decreased. There was less change in the percentage of cycles without luteal activity and E2 either within normal range or with peaks above 400 pg/ml.
FEBRUARY
1990 VOL. 41 NO. 2
173
CONTRACEPTION
TABLE IV PROPORTIONOF CYCLES WITH LUTEAL ACTIVITYAMONG NORPLANTRUSEPuS WITH MlRMAL !3STRADIOL YEARS OF USE
LUTEAL ACTIVITY
1st 2nd 3rd 4th 5th
2/8 2/6 8/15 8/12 9/12
25% 33% 53% 67% 75%
CONTROLS
14/14
100%
E2 NORMAL E2 NORMAL / P ) 14%
/ P ) 3
2 LOW I P ( 3 27% E2 LOW I P ( 3 2%
E2 NORMAL 25%
/ P <3 E2 ) 400 / P ( 3 27%
~4OO,P~3 32%
YEAR
1 AND
2
YEAR
3, 4, AND 5
Figure 1: Ovarian endocrine rofiles observedduring differentperiods of use of NORPLANTa subdernal implants.
174
FEBRUARY 1990 VOL. 41 NO. 2
CONTRACEPTION
DISUJSSIDN The most obvious advantage of continuous low-dose progestagen contraceptionover other hormonal methods is the smaller amount of exogenous hormones received by the users. The expectation is to obtain the same efficacy without the side effects which occur with higher doses. In fact, the history of hormonal contraceptionhas been characterizedby the progressive The low doses of reduction in the amount of administered hormones. progestagensand estrogensin the more recent pill formulations,have resulted in fewer side effects and lower risks of severe complications. A method with even lower daily dose and circulating levels of progestagens,without any estrogen,would be expectedto have even less health risks for the users. The effect of the exogenous hormones on circulatingsteroids is at least part of the intended contraceptiveeffect. The main differencebetween the continuous low-dose progestagen-onlysubdermal implantsand the combinedpill or depot-injectables,is that the first is not necessarilyanovulatory,but still is as effectiveor even more effectivethan the latter. Sev ral studies have shown endocrinological changes during use of NORPLAN?!. High and broad E peaks, anovulationand luteal insufficie y have all been describedas part oz the ovarian endocrinepicture in NORPLAN? users (2,3,4,5,6,7,8).What was still lacking was a more precisedescriptionof the quality, frequencyand dimensionof the changes and their variationalong time of use. The present study shows that, as expected,not all NORPLAFI?users have the same ovarian endocrine response to the continuous exposure to levonorgestrel. The observed effect varies from an apparently profound depressionof ovarian function,with persistentlylow E2 levels without luteal activity in some cycles, in some women, to practicallyno change as compared to normal cyclic E2 and P in other women. In one-thirdof the cycles studied, luteal activitywas observed. Length of use has a clear influenceon luteal function. Not only the proportionof subjects with luteal activityincreaseswith time but higher P levels are more frequentlyobserved in the fourth and fifth year of use. On the other hand, low E2 levels foE periods of 30 or more days, had already been found among users of NORPLANT (3,6). Accordingto our results, this happens in about one out of every four cycles during the first two years of use, in one out of 20 in the third year of use, and is not observed thereafter. However, the method used in this study, e.g. 10 samples in a 5-week period, is acceptable to determine the proportion of cycles with luteal activity but not necessarilyto define persistentlylow E2 secretion. We observed only a relativelysmall "window" of the general ovarian function:a different view could be obtainedby prolongingthe sampling for more weeks. Our results illustratethe effect of length of samplingover the apparent findings: in addition to the 11 subjects with E2 below 75 pg/ml in all samples, we had 12 with the same low levels in 8 or 9 of the 10 samples, meaning that if we had limited the sampling to 4 weeks, we would have artificiallydoubled the proportionwith low E2 levels.
FEBRUARY 1990 VOL. 41 NO. 2
175
CONTRACEPTION
Similarly, Croxatto et al. (6) found only one out of 61 cycles studied with all E2 levels below 100 pg/ml., but some of their samplingruns covered six weeks of observation instead of five. Their study included a smaller number of cycles within the first years of use (11/61 as comparedto 37/92 in our study), which may also contribute to explain the difference with our results. Presently,persistentlylow E2 levels (below40 &ml) are automatically linked to a higher risk of osteoporosis. Only three of our 92 observed cycles maintained levels below 40 pg/ml during the sampling period. The weight of these three patients was very low (41, 43 and 54 Kg), thereforethese women probably had greater ovarian su@oressiondue to higher levonorgestrellevels, as has been observed in NORPLANT users with low body weight (1). The question as to how long these low E2 levels are in fact maintained, would be answered only with a much more prolonged study, with all the practical problems it would present. However, the real dimensionof such risk may be reduced or eliminatedin these women since they are under continuous progestin administration,which has ,been shown to have a protectiveeffect against bone loss (9, 10). The presence of occasionalE peaks, prolonged for up to 10 days, could also seem alarming, considering w e absence of luteal activity in all those cycles. It does not, however,constitutea situationof unopposedestrogenic activity as observed in other clinical confflitions, as for example polycystic ovary syndrome. In the case of NORPLANT users, the subject's tissues are simultaneously under the continuous effect of a progestagen with potent antiestrogeniceffect. Moreover, we did not find any period longer than 10 days (three samples)with E2 over 400 pg/ml. In continuous low-dose, progestagen-only summary, women u@ng contraceptionthrough NORPLANT , have a variabledegree of endocrineactivity as compared with the complete depression of ovarian function among pill or injectables'users. Ovarian activity&ecomes closer to normal during the 3rd through 5th year of use of NORPLANT . This lesser depression of ovarian endocrine function explains the rapid resumption of ovulation (11) and recovery of fertilityafter discontinuationof use by removalof the implants (12, 13). Acknowledgements: This study was undertaken as part of the contraceptivedevelopment program sponsored and coordinated by the International Committee for ContraceptionResearch of the Population Council, New York. The financial support provided by the US Agency for International Development, the RockefellerFoundation,the George Hecht Fund, the Andrew W. Mellon Foundation and the United Nations Fund for Population Activities is gratefully acknowledged.
176
FEBRUARY 1990 VOL. 41 NO. 2
CONTRACEPTION
REFERENCES 1.
Sivin I. International Experience with NORPLANTR and NORPLANTContraceptives. Studies in Family Planning19:81-94 (1988).
2.
Weiner E, Johansson EDB. Plasma Levels of d-Norgestrel,Estradiol and Progesterone during Treatment with Silastic Implants Containing d-Norgestrel. Contraception14:81-92 (1976).
3.
Bleeding and Serum Moore DE, Roy s, Stanczyk FZ, Mishell DR. d-Norgestrel, Estradiol and Progesterone Patterns in Women Using d-Norgestrel Subdennal Polysiloxane Capsules for Contraception. Contraception17:315-328(1978).
4.
Shaaban MM, El-NasharIM, GhaneimahSA, Gomaa AA, Salah M, Abdel-AleemA. Hormonal Changes ring the First Year of Use of SubdermalLevonorgestrel Implants,NORPLAN? . Contraception30:391-405(1984).
5.
Croxatto HB, Diaz S, Paves M, Miranda P, BrandeisA. Plasma Progesterone Levels During Long-term Treatment with LevonorgestrelSilastic Implants. Acta Endocrinol.101:307-311(1982).
6.
Croxatto HB, Diaz S, Pavez M, Brand 's A. EstradiolPlasma Levels During Long-Term Treatment with NDRPLAN? Subdermal Implants. Contraception 38:465-475(1988).
7.
Brache V, Faundes A, Johansson E, Alvarez F. Anovulation, Inadequate Luteal Phase and Poor Sperm Penetration in Cervical Mucus During ProlongedUse of NORPLANTRImplants. Contraception31:261-273(1985).
8.
Alvarez F, &ache V, Tejada AS, Faundes A. Abnormal Endocrine Profile Among Wmen with Confirmed or Presumed Ovulation During Long-Term NORPLANTa Use. Contraception33:111-119(1986).
9.
Lindsay R, Hart DM, Purdie D et al. ComparativeEffects of Estrogen and Progesterone on Bone Loss in PostmenopausalWomen. Clin Sci Mol Med 54:193 (1978).
Effect of 10. Davidson BJ, Deftos LJ, Meldrum DR, Judd HL. MedroxyprogesteroneAcetate (MPA) on Bone Metabolism of Postmenopausal Women. Abstr No.279, Society for GynecolcgicInvestigation,27th Annual Meeting, Denver,CO, March 1980. Ovulation Detection 11. Ismail AAA, Anwar MY, Youssef SM, Toppozada M. Following Removal of LevonorgestrelSubdermal Contraceptive Implants. Contraception35:207-214(1987). 12. Affandi B, santoso SSI, Dapjadilaga HadisaputraW et al. Pregnancy After Removal of NORPLANT Implants Contraceptive. Contraception 36:203-209(1987). 13. Diaz S, Paves M, Cardenas H, Croxatto HB. Recovery of Fertility and Outcome of Planned PregnanciesAfter the Removal of NORPLANTR Subdermal Implants or Copper-T IUDs. Contraception35:569-579(1987).
FEBRUARY 1990 VOL. 41 NO. 2
177