Facts about an implantable contraceptive: Memorandum from a WHO Meeting*

Bulletin ofthe World Health Organization, 63 (3): 485-494 (1985) © World Health Organizationi 1985 Facts about an implantable contraceptive: Memorandum from a WHO Meeting* This Memorandum reviews the results of research undertaken in animals and human subjects on the implantable contraceptive, Norplant, and where relevant, its components, levonorgestrel and Silastic. Results from clinical trials, including effectiveness and side- effects, are evaluated and service delivery aspects commented on. The Memorandum concludes with a statement regarding the use of Norplant as an option for long-term reversible contraception. Through its Special Programme of Research, Development, and Research Training in Human Reproduction, the World Health Organization provides advice to Member States and responds to questions from governments, other organizations of the United Nations system, and the scientific com- munity on the safety and effectiveness of existing and new methods of fertility regulation. In response to a request from the United Nations Fund for Population Activities (UNFPA) to evaluate proposals for the introduction into family planning programmes of the new implantable contraceptive, Norplant, a consultation was convened by the Special Programme to review all data available from animal and human studies on this method of fertility regulation. The only implantable contraceptive that can currently be considered for introduction into family planning programmes is Norplant, which was devel- oped by the Population Council, New York, USA, primarily through its International Committee for Contraception Research (ICCR). It is registered under the Population Council's trademark as a subdermal implant for the release of levonorgestrel and is manufactured under licence from the Population Council by Huhtamaki Oy/Leiras Pharmaceuticals, Finland. It was approved for con- traceptive use in November 1983 by the Finnish National Board of Health and is presently under con- sideration by the Swedish National Board of Health and Welfare. Submissions are expected to be made to the drug regulatory agencies of some 40 additional countries within the next two years. About 20 years ago it was shown that Silastic (poly- dimethylsiloxane) could be used for the long-term * This Memorandum was drafted by the signatories listed on p.492, on the occasion of a Consultation convened by the World Health Organization in Geneva in October 1984. Requests for reprints should be addressed to the Director, Special Programme of Research, Development and Research Training in Human Repro- duction, World Health Organization, 1211 Geneva 27, Switzerland. A French translation of this Memorandum will appear in a later issue of the Bulletin. delivery of lipophilic drugs (1), and it was subse- quently proposed for use in the delivery of contra- ceptive steroids (2). After studies with several steroids (3-7), the ICCR selected levonorgestrel as- the drug with which the Silastic subdermal implant would be developed further (6, 8). Levonorgestrel is used exten- sively as a progestogen for contraception in both combined progestogen-estrogen and progestogen only oral contraceptive pills (9) and in other injectable and implantable systems, intrauterine devices, and vaginal rings currently under development (10). Since 1975, clinical studies on Norplant have been initiated in several centres by ICCR. In more recent years, other investigators, funded by other agencies, have also become involved. To date, studies have been undertaken in the following countries: Brazil, Chile, Colombia, Denmark, Dominican Republic, Ecuador, Egypt, Finland, India, Indonesia, Jamaica, Sweden, Thailand, and the United States of America. The Norplant system consists of six capsules, each containing 36 mg of levonorgestrel and having a diameter of 2.4 mm and a length of 3.4 cm. The six capsules appear to release levonorgestrel at a rate of approximately 80 ug per 24 h during the first 6-18 months of use. This rate declines over the next few months and thereafter the capsules deliver approxi- mately 30 ,g of levonorgestrel per 24 h. This latter rate of release is maintained for at least 5 years. The capsules are inserted subdermally, usually in the inner aspect of the upper arm or in the palmar aspect of the forearm, by means of a specially designed 10-gauge trocar which is introduced through a single 2-mm incision. A second-generation system is also being developed by the Population Council which utilizes only two implants and is known as Norplant-2. This system consists of two rods in which levonorgestrel is homogeneously dispersed within a Silastic matrix and 4545 -485- MEMORANDUM covered by a layer of Silastic. The rods are 2.4 mm in diameter and 4.4 cm long and have been shown to deliver amounts of levonorgestrel equivalent to those released by Norplant. Norplant-2 is currently being tested in phase III clinical trials. The only other implantable system for contra- ception that could be available to family planning programmes this decade is the Capronor device, developed by the Contraceptive Development Branch of the US National Institute of Child Health and Human Development in collaboration with theWHO Special Programme of Research, Development, and Research Training in Human Reproduction (10). Capronor consists of a single biodegradable capsule that delivers levonorgestrel for a period of approxi- mately 18 months. This system is at an early stage of development; preliminary clinical trials have been completed (11, 12) and studies will shortly be expanded to cover safety and efficacy. ANIMAL STUDIES The Toxicology Review Panel of the Special Programme has reviewed the animal data available on the two components of the system, levonorgestrel and Silastic, as well as on Norplant itself. Toxicological studies on levonorgestrel have been conducted in accordance with the requirements of the United Kingdom Committee on the Safety of Medicines and the United States Food and Drug Administration (FDA), among others. It should be noted that these studies refer to both the racemic mixture, d,l-nor- gestrel and to its biologically active enantiomer, levonorgestrel, which comprises 50Gb of the racemic mixture. Originally, d,l-norgestrel was used in hormonal contraceptives but in more recent times it has largely been replaced by levonorgestrel. Acute toxicity The acute toxicity of d,l-norgestrel is over 5000 mg/kg of body weight (LD5o) as determined by oral administration in dogs and by various routes of administration in adult mice and rats. The LD50 value is 950 mg/kg of body weight in neonatal mice and more than 1260 mg/kg of body weight in neonatal rats. The dose administered to women by implants may be estimated to be about 1 gg/kg of body weight per day. No toxicity relevant to the human was seen in animal studies, which indicates a wide margin of safety. Subacute and chronic toxicity A considerable amount of information is available from long-term studies of d,I-norgestrel and levonor- gestrel in mice, rats, dogs, and monkeys. The chronic toxicity studies with levonorgestrel, in dogs and monkeys, are summarized in Table 1. These studies, conducted by Wyeth Laboratories, USA, were reported in detail to the FDA, and have been incor- porated in files on Norplant submitted by the Population Council to the FDA, and by Huhtamaki Oy/Leiras Pharmaceuticals to the Finnish National Board of Health. The Toxicology Review Panel noted that studies in the beagle dog showed no adverse effects relevant to human use, but nevertheless restated its opinion that the beagle dog is an unsuitable toxicological model for the study of progestogens. The studies were completed before the Panel's opinion on the relevance of the beagle model was first published in 1982 (13). One 10-year study in the rhesus monkey, in which the animals received the drug orally in capsules, has recently been completed. The highest dose given was equivalent to 1000 times the human implant dose. No toxic effects attributable to the treatment were found. Extensive chronic toxicological studies have been conducted to evaluate the systemic effects of levonor- gestrel. Dosing schedules have included daily administration, as well as cyclic administration comparable to the human use of contraceptives. The studies have served as a basis for the approval by the FDA of oral contraceptives with the following composition: 500 ug of d,l-norgestrel with 50 1tg of ethinylestradiol; 300 ug of d,l-norgestrel with 30 ug of ethinylestradiol; 150 jg of levonorgestrel with 30 jg of ethinylestradiol; and 75 jg of d,l-norgestrel alone. Table 1. Chronic toxicity studies on levonorgestrel Species Highest dose Route Multiple of Duration (mg/kg of body weight) daily implant dose Dog 0.5. cyclic Capsules 500 7 years Dog 0.125, cyclic Capsules 125 7 years Monkey 1 .0, cyclic Capsules 1000 10 years 486 AN IMPLANTABLE CONTRACEPTIVE Effects on reproduction Teratological and reproduction studies have been undertaken in rats and rabbits. Both d,l-norgestrel and levonorgestrel have been used in doses corres- ponding to more than 100 times and 50 times the human dose, respectively. No adverse effects that would suggest a risk for humans were found. In another study, in which covered rods releasing 3.5 Ag of levonorgestrel per 24 h were inserted between implantation sites in the uterus of the pregnant rabbit, there was no evidence of any teratological effects on the progeny (ICCR, unpublished data). Masculini- zation was observed in offspring of rats, but this has been reported with other progestogens and is not considered to be relevant to human subjects. No other fetal abnormalities or malformations were found and the duration of gestation was not altered. Offspring reserved for subsequent breeding were normal with respect to survival age and reproductive capacity. Three dose levels of d,l-norgestrel were given to lactating rats, the highest dose being 150 times the human dose. No significant effect was seen in the mothers, in the young, or on lactation. Relevance of animal tests to the implant dosage form Long-term animal tests of levonorgestrel have been carried out using oral administration. Several factors must be considered in weighing the relevance of these tests to the implant mode of administration. These include the relative bioavailability of the drug by different routes of administration and the relative constancy of exposure to the drug. When plasma levels of the drug have been measured, the bio- availability by the oral route has been reported as 9% for the rat, 22% for the dog, 9% for the monkey, but 87% for the human (14). Plasma levels of levonor- gestrel in Norplant users are lower than those observed with oral administration and do not fluc- tuate. In view of these considerations the toxico- logical studies involving oral administration were considered pertinent to the safety of the drug released from the implanted drug delivery system. Safety ofsilicone rubber in implants Silicone rubber materials have been used to make implants of various kinds, such as heart valves and drainage tubes, for use in humans over long periods. It was estimated that, by 1974, 200 000 shunts had been implanted in patients with hydrocephalus, and many thousands more have been used in the subse- quent 10 years. The Silastic tubing used to make the shunts is of the same type and diameter as that used in the Norplant capsules. Even though it is well established that impervious or semipermeable subcutaneous implants induce local sarcomas in rats (15), this observation is not considered to be relevant to man in the light of the very extensive and long-term use of Silastic in humans, during which malignant tumours have not been observed. Conclusions The Toxicology Review Panel concluded that the toxicological and teratological data from animal studies on both levonorgestrel and Silastic provide a sufficient indication that Norplant may be considered safe for use in humans. HUMAN STUDIES A substantial number of clinical trials have been undertaken on Norplant, including several major phase III studies supported by the ICCR and other agencies. Use-effectiveness and side-effects have been evaluated in studies that have continued for more than 5 years. However, because the method is only now becoming widely available, there have, as yet, been no epidemiological studies of long-term safety. Release rates The amount of levonorgestrel released from implants has been calculated from the difference between the steroid content at time of placement and the amount remaining after a known period of use. An analysis of steroid loss from 42 sets of implants used for periods ranging from 1.3 to 6.3 years gave an average release rate of 30 yg per 24 h (16). These data were later supplemented by similar analyses from an additional 117 sets of capsules recovered at various time intervals up to about 8 years from subjects in several clinics. The data suggest an initial release rate of approximately 80 zg per 24 h, which declines over the first 6-18 months to an approximately constant release rate of about 30 tg per 24 h. Peripheral blood levels of levonorgestrel Several studies have reported data on peripheral blood levels of levonorgestrel (16-21). Extensive data from unpublished studies were also made available to the consultation by the ICCR. In all of the studies, analyses for levonorgestrel were carried out by radio- immunoassay. The results show that immediately after implantation of the devices there is a rapid rise in peripheral levonorgestrel levels, to 3-4.5 nmol/litre after 24 hours. From results derived from other 487 MEMORANDUM studies on levonorgestrel, levels of this magnitude are likely to be more than sufficient to inhibit ovulation (22, 29). During the subsequent 6-18 months, the mean plasma levels gradually decline to 1-1.5 nmol/litre. The mean values in individual subjects appear to remain constant thereafter for a further 5 years, after which there is a substantial decrease. As with all other constant delivery systems for levonorgestrel, there is considerable between-subject variation in circulating levels of levonorgestrel. One of the factors that contribute to this variation is the large difference in sex-hormone binding globulin (SHBG) levels observed between normal individuals (23, 26). Unlike several other progestogens, levo- norgestrel has a high binding affinity to SHBG, and there is thus a correlation between circulating levels of SHBG and levonorgestrel. Levonorgestrel also suppresses SHBG levels, resulting in a decrease in both SHBG and levonorgestrel levels. Moreover, subjects starting with a high level of SHBG show a greater suppression in SHBG and hence levonor- gestrel than those with lower initial levels of SHBG. Thus significant between-subject variation in peripheral levonorgestrel levels is found (24-26). Two studies on Norplant users have shown similar results (27; ICCR, unpublished data). Following the removal of the implants, plasma levonorgestrel becomes unmeasurable within a few days. A removal half-life of some 18 hours can be calculated (ICCR, unpublished data), which is in good agreement with that reported by other investi- gators for other modes of administration of levonor- gestrel (17, 28, 29). Effects on ovarian function The ovarian response to levonorgestrel released from the Norplant implants shows considerable vari- ation, both within and between subjects, as has been observed with other low-level levonorgestrel delivery systems (29). When periods of treatment were evaluated in 6-week segments, elevated plasma pro- gesterone levels ( > 10 nmol/litre) were observed in 4401 of the segments, but the mean increase in pro- gesterone level was less than that observed in ovula- tory cycles in normal menstruating women. There was also a marked suppression of the luteinizing hormone surge. It is not known whether the periodic increases in progesterone levels seen in Norplant users and in subjects using other low-dose progestogen formu- lations (29) reflect the occurrence of ovulation, since monitoring by ultrasound has as yet yielded incon- clusive results. In other treatment segments, increases in estradiol were observed, which varied in amplitude (up to 2 nmol/litre) and duration (5-20 days), in the absence of elevated progesterone levels, indicating follicular activity and no ovulation (17, 18). Effects on cervical mucus The effect of levonorgestrel on cervical mucus also contributes to its contraceptive efficacy. During Norplant use, the cervical mucus becomes viscous and scanty, and in a few subjects from whom it was pos- sible to obtain cervical mucus, poor sperm pen- etration was seen compared with control subjects. In another study, post-coital tests have also shown poor sperm penetration (ICCR, unpublished data). Effects on the endometrium Endometrial biopsies from 41 women who used Norplant for between 6 and 110 weeks have been studied (30). Using the criteria of Noyes et al. (31), 5 endometria were normal, 12 were diagnosed as irregular secretory, and 24 demonstrated suppressed activity. These findings are similar to those obtained with other low-dose progestogen delivery systems (29). There was no evidence of endometrial inflam- mation, atrophy, hyperplasia, carcinoma, or other pathological processes. Other endocrine effects Although one study claimed a significant increase in peripheral androstenedione and some increase in testosterone levels after 6 months' use of Norplant (32), two other studies demonstrated contrary results (27, 33). One of these, a cross-sectional study, showed significantly lower testosterone levels in users of Norplant after 51 months than in IUD users (33), while the other, a longitudinal study, showed a decline in testosterone, androstenedione, and SHBG levels (27). Increases in prolactin levels have been observed, but were within normal limits (34). With regard to effects on adrenal function, a decrease in peripheral levels of cortisol, of approxi- mately 30% over 12 months, has been observed, although the values were within the normal range and remained so during 3 years of observation (35). Similar results were obtained in another study over one year (36). The adrenal response to adrenocortico- tropic hormone (ACTH) stimulation has been shown to be normal (37). Regarding thyroid function, no alterations have been observed in thyroxine and thyroid-stimulating hormone (TSH) levels; a decrease was found in triiodothyronine levels, although the levels remained within the normal range (33). Thus, Norplant does not appear to disturb significantly the function of the thyroid and adrenal glands. 488 AN IMPLANTABLE CONTRACEPTIVE Metabolic effects Several cross-sectional and longitudinal studies have been undertaken to determine the metabolic and biochemical effects of Norplant at intervals of up to 3 years. Studies on lipid metabolism have given inconsistent results: triglycerides were reported to be significantly decreased, as were cholesterol and low-density lipo- protein (LDL)-cholesterol levels, with up to 37 months of use (33, 38, 39; ICCR, unpublished data). Widely discrepant changes in high-density lipoprotein (HDL)-cholesterol have been reported, however, including: no change (33); decrease (39); and no change initially, but a significant increase at 12 months (38). Two laboratories reported decreases in HDL-cholesterol levels and in total cholesterol, giving an unchanged cholesterol/HDL-cholesterol ratio (ICCR, unpublished data). In the centre reporting increased HDL-cholesterol levels, the cholesterol/HDL-cholesterol ratio was reduced at 6 months and further at 12 months. A recent study with rods releasing equivalent amounts of levonorgestrel has shown alterations in some subfractions of lipo- proteins; HDL-2a, HDL-2b, and HDL-3 were reduced after 1-2 months of use; these decreases were less pronounced at 6 months and there was a return to normal levels after 12 months (40). The studies carried out so far used different methodologies, tech- niques of quality control, and populations, and further studies are still needed to clarify the effects of levonorgestrel released from Norplant on lipids and lipoproteins. Liver function has been evaluated at several centres (41; ICCR, unpublished data). Slight non-significant increases in total serum protein and albumin were observed. A reduction in ceruloplasmin levels at 6 months was reported, as well as small increases in bilirubin and bile acids in the first month; these levels returned to normal over the next 5 months (41). No other changes in liver function were seen. In two other cross-sectional studies (33, 42), after 51 months of use, liver function was not significantly different from that observed in control subjects using intra- uterine devices. Measurements of urea nitrogen, uric acid, sodium, potassium and calcium (33, 42; ICCR, unpublished data) showed no changes. A significant decrease in inorganic phosphorus was observed, but levels remained within the normal range (33). No changes in immunoglobulins have been observed. Mean blood glucose levels have been shown to be elevated, but within the normal range, after 12-51 months of Norplant use (ICCR, unpublished data). In one study, oral glucose tolerance tests were done at intervals of 6 months for 3 years in 9 subjects (35). Abnormal glucose tolerance was observed in 2 women with familial diabetes and, in both, glucose tolerance returned to normal at the next test. While the data are limited, Norplant appears to have no deleterious effect on carbohydrate metabolism. Coagulation studies before treatment, and at 1, 3, and 6 months were undertaken in one study (43). Two groups of subjects taking oral contraceptives were used as controls. Nine clotting factors and nine other haematological parameters were measured. An increase in factor VII and a decrease in antithrombin III were the only significant changes observed in Norplant users after 6 months; these changes were significantly less than those observed in the groups taking oral contraceptives. Moreover, the oral contraceptive users showed significant changes in many of the other factors measured, whereas the Norplant users did not. Use-effectiveness and discontinuation rates Experience with more than 14 000 insertions of Norplant has shown a low pregnancy rate and a high continuation rate (44-48; ICCR, unpublished data). The aggregate interim results from a number of clinical trials conducted by the ICCR, involving 992 women, indicate a gross cumulative pregnancy rate at 5 years of 2.6 per 100 woman-years. The annual pregnancy rates during the first 5 years ranged from 0.2 to 1.3. Data on 100 women entering the 6th year show a pregnancy rate of 2.9 (with a standard error of± 1.4), which suggests a reduction in efficacy although the rate is not significantly different from the rates for the first 5 years. In a clinical trial in 3 countries conducted between the years 1978 and 1983, the continuation rate at 1 year was 81.4 per 100 women and at 4 years was 49.1 per 100 women (44). In a number of other studies undertaken in developing countries, continuation rates at 1 year were higher (45-48). Continuation rates were lowest among younger women and those desiring a pregnancy (44). These continuation rates are comparable to those commonly quoted for the IUD. The major medical reason for termination has been menstrual irregularities, the aggregate cumulative rate for several centres being 14.1 per 100 women at 4 years (44). It should be noted that these data are from clinical trials and therefore may not be representative of discontinuation rates that might be seen in pro- gramme use. It is difficult to compare the results of these studies with those on other contraceptive methods since the contraceptive in question has a known duration of action of 5 years and the acceptors are thus likely to be highly motivated to continue contraceptive use for such a period. A total of 11 pregnancies have been reported to be due to method failure. Of these, there were 2 ectopic 489 MEMORANDUM pregnancies, both from the same centre, giving an incidence of ectopic pregnancy of 0.8 per 1000 woman-years (44). Although the number of events is small, the method does not appear to increase the incidence of ectopic pregnancy. However, should a pregnancy occur in a woman using Norplant, atten- tion must be paid to the possibility of an ectopic pregnancy. In a cohort of 992 women, a mean increase in weight of 1.4 kg occurred during the first year of use of Norplant. However, although weight was increased in 52% of the women, 35% of the subjects showed a decrease (49). No significant changes in systolic or diastolic blood pressure were found during the first 3 years of use of Norplant; no data are available for a longer duration of use. In the studies reported by ICCR, headache was the second pre- dominant spontaneously reported complaint after bleeding disturbances, accounting for 3.2% of the discontinuations during the first 3 years of use (44). However, it has not appeared to be a problem in most clinical trials of Norplant in other countries (45-48). Bleeding patterns As with other progestogen-only contraceptives, the major side-effect of Norplant is the disruption of the menstrual cycle; about 40% of Norplant users report a regular bleeding pattern during the first year of use. Most abnormal bleeding episodes are characterized by an increased frequency, an increased number of bleeding days, or irregular bleeding and spotting. Heavy bleeding episodes are uncommon and the prevalence of abnormal bleeding decreases with duration of use. In a study of a cohort of 104 Norplant users who completed 4 years of use, 25% reported bleeding episodes lasting more than 11 days in the first year, while only 10% had such bleeding episodes during the second, third and fourth years of use (44). In another cohort of 54 women, 50Gb of users reported bleeding and spotting for more than 96 days during the first year of use; only 20% reported this during the third year (50). While the menstrual problems do diminish markedly with time, these complaints are the major reason for discontinuation in the first year of use (45, 48,51,52). In a small study, the mean amount of monthly blood loss in Norplant users decreased, although not significantly, from 30 ml pretreatment to 24 ml during the first year of use (53). Mean haemoglobin levels increased slightly but significantly in Norplant users and evidence of anaemia from the use of Norplant, even in women with frequent bleeding episodes, has been rare (49). If a user develops severe and prolonged bleeding, removal of the capsules should be considered. Normal regular menstrual bleeding can be expected to resume shortly after, unless the bleeding has another pathogenesis. Insertion and removal ofNorplant Extensive clinical experience has been accumulated on the insertion and removal of Norplant. More than 14 000 women have had these capsules inserted and 10-15% have had them removed. The experience gained has indicated that, to minimize the risk of infection, both insertion and removal should be performed in a clinic setting. It is of utmost importance that sterile techniques be maintained throughout both procedures. Norplant should be inserted during the first 7 days after the onset of menstruation in order to minimize the risk of insertion during an undiagnosed preg- nancy. Insertion can also be done immediately after abortion and in non-breast-feeding women im- mediately postpartum. At this stage of knowledge, it is not possible to make a recommendation on the use of Norplant in lactating women. The preferred sites of insertion are the upper arm or the forearm. It is important that the capsules be inserted superficially, as deeper insertion makes removal more difficult. In one study, removal was difficult in 1201 of the subjects primarily because the capsules had been inserted too deeply (ICCR, unpublished data). Sutures are not required to close the small incision and the scar is usually minimal, although keloid formation has been reported to occur occasionally (ICCR, unpublished data). Local complications following insertion are uncommon and the cumulative local infection rate is 0.3%o. Expulsion of capsules is rare and has occurred only in the presence of infection. Only one instance of migration of a capsule has been reported (ICCR, unpublished data). The capsules should be removed after 5 years of use as the effectiveness of the method declines after this time. However, in the event that a woman does not return for scheduled removal, no deleterious effects would be expected, other than the risk of unplanned pregnancy. If a woman wants to continue this method of contraception, a new set of implants may be inserted. Removal of the capsules is more time- consuming than the insertion: for example, in one clinic the mean time for insertion was 7 minutes, and for removal, 21 minutes (ICCR, unpublished data). When the capsules are removed because of compli- cations related to the method or if the woman wants to conceive, blood levels of levonorgestrel decline rapidly, becoming undetectable within two days, and ovulation resumes promptly. Of patients having Norplant removed in order to conceive, 40% had become pregnant by 3 months, 76% by 1 year and 490 AN IMPLANTABLE CONTRACEPTIVE 90070 by 24 months (44). These rates are similar to normal rates of fecundity. The implants should be removed whenever requested by the woman or in the case of an accidental pregnancy that the woman wishes to carry to term. Contraindications for the use of Norplant Since Norplant is a new method of contraception, there has been insufficient time for large-scale, long- term studies of rare adverse effects; thus, contraindi- cations and possible warnings regarding Norplant use must be based on extrapolation from information on other hormonal methods. Until further information is available it is considered inadvisable to use Norplant as a method of contraception in women undergoing anticoagulant therapy, or with undiag- nosed abnormal uterine bleeding, known or suspected pregnancy, haemorrhagic diathesis, or active hepato- cellular disease. SERVICE DELIVERY ASPECTS Service delivery issues arising from the intro- duction of Norplant must be addressed if utilization of this method in family planning programmes is to be successful. Although clinical trials have provided extensive experience with Norplant, there is limited information on service aspects, which will be needed by both family planning programme administrators and clinic staff. The cost-effectiveness of the method relative to other methods of contraception remains to be determined. Since Norplant utilizes a new mode of drug delivery, requiring insertion and later removal, considerable attention must be given to training for these procedures. Available information suggests that both physicians and other health personnel can easily be trained to insert and remove the capsules (51). In addition, clinic staff should be familiar with the bleeding disturbances that are particularly common during the first year, and need to be trained to counsel users accordingly. It is important to identify specific centres that can serve as national or regional resources for the training of health personnel. Similarly, refer- ral or back-up facilities need to be identified, especially in rural areas. From the outset, it is essential that government health agencies be involved in the introduction of the method. Experience with the introduction of other new methods of contraception has shown the essential role that "pre-introductory trials" play in this pro- cess. These trials can be of use in identifying training requirements and other method-related service delivery issues. CONCLUSIONS The Norplant subdermal implant is a very efficient method of contraception, which remains effective for at least 5 years. This long duration of contraceptive action is particularly advantageous to women who wish an extended period of contraceptive protection, irrespective of whether a subsequent pregnancy is planned. Animal studies on levonorgestrel, the progestogen released by the implant, provide a sufficient indi- cation that the drug may be considered safe for use in humans. Extensive clinical studies have so far not shown any major adverse side-effects. Silastic, from which the capsule is made, has been used in human subjects since 1950 for extended periods and has shown no adverse effects. The clinical trial data available to date on more than 4000 women have shown high continuation rates, ranging from 60%7o to 950% at the end of the first year, and about 50070 at the end of the fifth year. As with other progestogen-only contraceptives, the most common side-effect and the principal reason for discontinuation is bleeding irregularities. These occur at some time in the majority of women using this method, although bleeding problems tend to diminish with increasing duration of use. Women have reported irregular bleeding, spotting and, to a lesser extent, amenorrhoea. Heavy and prolonged bleeding has rarely occurred. Because of the constant release of small amounts of the contraceptive drug (approximately 30 jig per 24 h after the first 6-18 months of use), the total amount of drug to which the body is exposed is small, being less than 100 mg of levonorgestrel over 5 years. Thus, potential side-effects of the contraceptive steroid are minimized and there appears to be little significant effect on metabolic or other physiological param- eters. Furthermore, exposure to the drug can be terminated at any time by removal of the device. This modality is just beginning to be widely introduced, thus little information has been generated on rare or long-term effects. It is recommended that appropriate surveillance activities be developed to evaluate its long-term safety. As with other hormonal contraceptive methods, additional research is also needed on the use of Norplant during lactation, since few data are available on the effect of levonorgestrel on the growth and development of children exposed to the steroid via breast milk. The insertion and removal of Norplant, whether by trained physicians or by other health personnel, require appropriate clinic facilities. It is essential that adequate training programmes be established for clinic personnel prior to introduction of Norplant into a family planning programme. 491 492 MEMORANDUM "Pre-introductory" trials, prior to widespread introduction, are important for providing training and information about service delivery issues, and to assess the local acceptability of the method. it is important to note that the skill of the provider at insertion will largely determine the ease of removal. In summary, the participants in the Consultation concluded that Norplant provides an effective and reversible long-term method of fertility regulation. It is considered suitable for use in family planning pro- grammes, along with other currently available contra- ceptive preparations and devices, since it provides an important option for women desiring long-term contraception. * * * K. Adadevoh, Ebute Metta, Lagos, Nigeria B. Affandi, University of Indonesia, Jakarta, Indonesia U. Borell, Karolinska Institute, Stockholm, Sweden E. Diczfalusy, Karolinska Institute, Stockholm, Sweden K. W. von Eickstedt, Bundesgesundheitamt, Berlin (West) M. Elder, University of London, London, England R. Gray, Johns Hopkins University, Baltimore, MD, USA S. Koetsawang, Mahidol University, Bangkok, Thailand Suman Mehta, Indian Council of Medical Research, New Delhi, India D. Mishell, University of Southern California, Los Angeles, CA, USA S. Ratnam, University of Singapore, Singapore Pramilla Senanayake, International Planned Parenthood Federation, London, England M. Shaaban, University of Assiut, Assiut, Egypt R. Shearman, University of Sydney, Sydney, Australia (Chairman) K. H. Yang, Korea Advanced Institute of Science and Technology, Seoul, Republic of Korea Representatives from the Population Council, New York, USA W. Bardin G. Brown F. Greenslade H. Nash D. Robertson Representatives from Huhtamaki Oy/Leiras Pharmaceuticals, Finland H. Allonen H. Voipio S. Jarvela Representative from Finnida, Helsinki, Finland U.-M. Finskas Representative from UNFPA H. Corvalan, New Delhi, India Secretariat (World Health Organization, Geneva, Switzerland) P. E. Hall, Special Programme of Research, Development, and Research Training in Human Reproduction (Secretary) P. Corfman, Special Programme of Research, Development, and Research Training in Human Reproduction S. E. Holck, Special Programme of Research, Development, and Research Training in Human Reproduction A. Mundigo, Special Programme of Research, Development, and Research Training in Human Reproduction P. J. Rowe, Special Programme of Research, Development, and Research Training in Human Reproduction M. Ten Ham, Pharmaceuticals REFERENCES 1. FOLKMAN, J. & LONG, D. M. The use of silicone rubber as a carrier for prolonged drug therapy. Journal ofsurgical research, 4: 139-142 (1964). 2. SEGAL, S. J. & CROXATTO, H. B. Single administra- tion of hormones for long-term control of reproductive function. Presented at the 23rd Meeting of the American Fertility Society, Washington, DC, 14-16 April 1967. AN IMPLANTABLE CONTRACEPTIVE 493 3. CROXATTO, H. B. ET AL. Fertility control in women with a progestin released in microquantities from sub- cutaneous capsules. American journal ofobstetrics and gynecology, 105: 1135-1139 (1969). 4. CROXATTO, H. B. ET AL. Contraceptive action of megestrol acetate implants in women. Contraception, 4: 155-159 (1971). 5. COUTINHO, E. M. ET AL. Long-term contraception by subcutaneous silastic capsules containing megestrol acetate. Contraception, 2: 313-317 (1970). 6. CROXATTO, H. B. ET AL. Clinical assessment of sub- dermal implants of megestrol acetate, d-norgestrel, and norethindrone as a long-term contraceptive in women. Contraception, 12: 615-619 (1975). 7. COUTINHO, E. B. & DA SILVA, A. R. One-year contra- ception with norgestrienone subdermal Silastic implants. Fertility and sterility, 25: 170-174 (1974). 8. SEGAL, S. J. Contraceptive subdermal implants. In: Mishell, D. R. Jr., ed., Advances in fertility research, New York, Raven Press, 1982, pp. 117-127. 9. REYNOLDS, J. E. F., ed., Extra pharmacopoeia, Martindale, 28th ed. London, Pharmaceutical Press, 1982 (levonorgestrel, pp. 1424-25; oral contraceptives, pp. 1402-06). 10. WORLD HEALTH ORGANIZATION, SPECIAL PROGRAMME OF RESEARCH, DEVELOPMENT AND RESEARCH TRAINING IN HUMAN REPRODUCTION. 12th Annual Report. Geneva, 1983, pp. 63-70. 11. ORY, S. J. ET AL. The effect of a biodegradable con- traceptive capsule (Capronor) containing levonorgestrel on gonadotropin, estrogen and progesterone levels. American journal of obstetrics and gynaecology, 145: 600-605 (1983). 12. WORLD HEALTH ORGANIZATION, SPECIAL PROGRAMME OF RESEARCH, DEVELOPMENT AND RESEARCH TRAINING IN HUMAN REPRODUCTION. 13th Annual Report, Geneva, 1984 (in press). 13. Facts about injectable contraceptives: Memorandum from a WHO Meeting. Bulletin of the World Health Organization, 60: 199-210 (1982). 14. DUSTERBERG, B. ET AL. Terminal half-lives in plasma and bioavailability of norethisterone, levonorgestrel and cyproterone acetate and gestodene in rats, beagles and rhesus monkeys. Contraception, 24: 673-683 (1981). 15. BISCHOFF, F. & BRYSON, G. Special requirements for plastic devices. In: Briggs, M. H. & Diczfalusy, E., ed., Pharmacological models in contraceptive develop- ment, Stockholm, Karolinska Institute, 1974, p. 300. 16. DIAZ, S. ET AL. A five-year clinical trial of levonor- gestrel silastic implants (Norplant). Contraception, 25: 447-456 (1982). 17. WEINER, E. & JOHANSSON, E. D. B. Plasma levels of d-norgestrel, estradiol and progesterone during treat- ment with silastic implants containing d-norgestrel. Contraception, 14: 81-92 (1976). 18. MOORE, D. E. ET AL. Bleeding and serum d-norgestrel, estradiol and progesterone patterns in women using d- norgestrel subdermal polysiloxane capsules for contra- ception. Contraception, 17: 315-328 (1978). 19. CROXATTO, H. B. ET AL. Plasma levels of levonorgestrel in women during long term use of Norplant. Contra- ception, 23: 197-209 (1981). 20. ALVAREZ, F. ET AL. Levonorgestrel plasma levels during continuous administration with different models of sub- dermal implants. Contraception, 27: 123-130 (1983). 21. ROBERTSON, D. N. ET AL. Release rates of levonor- gestrel from silastic capsules, homogeneous rods and covered rods in humans. Contraception, 27: 483-495 (1983). 22. LANDGREN, B. M. ET AL. A pharmacokinetic and pharmacodynamic study of a new type of levo- norgestrel-releasing vaginal device. Contraception (1985) (in press). 23. EL MAKHZANGY, M. N. ET AL. Sex hormone binding globulin capacity as an index of estrogenicity or andro- genicity in women on oral contraceptive steroids. Clinical endocrinology, 10: 39-45 (1979). 24. VICTOR, A. ET AL. Relation between sex hormone binding globulin and d-norgestrel levels in plasma. Acta endocrinologica, 86: 430-436 (1977). 25. TODDYWALLA, V. ET AL. Pharmacodynamic effects of ethinyl estradiol in women using vaginal devices releasing small doses of levonorgestrel at a constant rate. Contraception, 28: 21-39 (1983). 26. CEKAN, S. Z. ET AL. The interaction between sex hormone binding globulin and levonorgestrel released from vaginal rings in women. Contraception, 41: 431 (1985). 27. OLSSON, S-E. ET AL. Androgen levels in women using Norplant implants. Contraception, (1985) (in press). 28. HUMPEL, M. ET AL. Investigations of pharmacokinetics of levonorgestrel to specific consideration of a possible first-pass effect in women. Contraception, 17: 207-220 (1978). 29. LANDGREN, B-M. ET AL. Pharmacokinetic and pharma- codynamic investigations with vaginal devices releasing levonorgestrel at a constant, near zero order rate. Contraception, 26: 567-585 (1982). 30. CROXATTO, H. D. ET AL. Histopathology of the endo- metrium during continuous use of levonorgestrel. In: Zatuchni, G. L. et al., ed., Long-acting contraceptive delivery systems, Philadelphia, Harper and Row, 1984, pp. 290-295. 31. NOYES, R. W. ET AL. Dating the endometrial biopsy. Fertility and sterility, 1: 3-25 (1950). 32. OSMAN, M. I. ET AL. Subdermal levonorgestrel implants: serum androgens. Contraceptive delivery systems, 4: 127-131 (1983). 33. CROXATTO, H. B. ET AL. Clinical chemistry in women treated with levonorgestrel implants (Norplant) or a TCu 200 IUD. Contraception, 27: 281-288 (1983). 34. SHAABAN, M. M. ET AL. Hormonal changes during the first year of use of subdermal levonorgestrel implants, Norplant. Contraception, 30: 391-405 (1984). 35. DIAZ, S. ET AL. A three-year clinical trial with levonor- gestrel silastic implants. Contraception, 19: 557-573 (1979). 36. BAYARD, M. A. ET AL. Serum cortisol in women users of subdermal levonorgestrel implants. Contraceptive delivery systems, 4: 133-135 (1983). 37. JOHANSSON, E. D. B. & ODLIND, V. Norplant: bio- chemical effects. In: Mishell, D. R., Jr., ed., Advances in human fertility and reproductive endocrinology, Vol. 2: Long-acting steroid contraception. New York, Raven Press, 1983, pp. 117-125. 494 MEMORANDUM 38. SHAABAN, M. M. ET AL. Effect of subdermal levonor- gestrel contraceptive implants, Norplant, on serum lipids. Contraception, 30: 413-419 (1984). 39. ROBERTSON, D. N. Norgestrel-releasing silastic rods: clinical effects, biochemical effects and in vivo release rates. In: Mishell, D. R. Jr., ed., Advances in human fertility and reproductive endocrinology, Vol. 2: Long- acting steroid contraception. New York, Raven Press, 1983, pp. 127-147. 40. Roy, S. ET AL. Long-term reversible contraception with levonorgestrel-releasing Silastic rods. Americanjournal of obstetrics and gynecology, 148: 1006-1013 (1984). 41. SHAABAN, M. M. ET AL. Effect of subdermal levonor- gestrel contraceptive implants, Norplant, on liver function. Contraception, 30: 407-412 (1984). 42. CROXATTO, H. B. ET AL. Clinical chemistry in women treated with progestogen implants. Contraception, 18: 441-450 (1978). 43. SHAABAN, M. M. ET AL. Effect of levonorgestrel contraceptive implants, Norplant, on blood coagulation. Contraception, 30: 421-430 (1984). 44. SIVIN, 1. ET AL. A four-year clinical study of Norplant implants. Studies in family planning, 14: 184-191 (1983). 45. SHAABAN, M. M. ET AL. A prospective study of Norplant implants and the TCu 380Ag IUD in Assiut, Egypt. Studies in family planning, 14: 163-169 (1983). 46. SATAYAPAN, S. ET AL. Perceptions and acceptability of Norplant implants in Thailand. Studies in family planning, 14: 170-176 (1983). 47. MARANGONI, P. ET AL. Norplant implants and the TCu 200 IUD: a comparative study in Ecuador. Studies in family planning, 14: 177-180 (1983). 48. LUBIS, F. ET AL. One-year experience with Norplant implants in Indonesia. Studies in family planning, 14: 181-183 (1983). 49. INTERNATIONAL COMMITTEE FOR CONTRACEPTION RESEARCH OF THE POPULATION COUNCIL. Contraception with long-acting subdermal implants. 2. Measured and perceived effects in international trials. Contraception, 18: 335-353 (1978). 50. SIVIN, 1. Clinical effects of Norplant subdermal implants for contraception. In: Mishell, D. R. Jr., ed., Advances in human fertility and reproductive endo- crinology, Vol. 2. Long-acting steroid contraception. New York, Raven Press, 1983, pp. 89-116. 51. AFFANDI, B. ET AL. Clinical trial of Norplant in Indonesia. Symposium on long-term subdermal contra- ceptive implants: Egyptian and international experience, 23-24 February 1984, Assiut, Egypt. 52. KOETSAWANG, S. ET AL. Norplant clinical study in Thailand. In: Zatuchni, G. I. et al., ed., Long-acting contraceptive delivery systems. Philadelphia, Harper & Row, 1984, pp. 459-470. 53. NILSSON, C. G. & HOLMA, P. Menstrual blood loss with contraceptive subdermal levonorgestrel implants. Fertility and sterility, 35: 304-306 (1981).