Cyclo-ProveraTM

International Journal of Gynecology & Obstetrics 62 Suppl. 1 Ž1998. S43]S56

New once-a-month injectable contraceptives, with particular reference to Cyclofem WrCyclo-ProveraTM P.E. HallU UNDPr UNFPAr WHOr World Bank Special Programme of Research, De¨ elopment and Research Training in Human Reproduction, World Health Organization, Gene¨ a, Switzerland

Abstract Once-a-month injectable contraceptives containing a progestogen and an estrogen have been developed that disrupt vaginal bleeding patterns less than the widely used progestogen-only preparations. Pharmacokinetic studies were undertaken of dosages and ratios of the progestogens and the respective estrogens. In Phase III clinical trials, annual pregnancy rates were below 0.4% for MesigynaW Žnorethisterone enanthaterestradiol valerate, Schering AG, Berlin, Germany. and below 0.2% for Cyclofem W ŽMPArE2C. Žmedroxyprogesterone acetaterestradiol cypionate, Aplicaciones Farmaceuticas, SA, Mexico and PT Tunggal, Indonesia.. More than two-thirds of women had predictable, regular cycles, and discontinuation due to bleeding-related problems occurred less than half as often as with progestogen-only injectables. With MPArE2C, return to fertility is similar to that observed with other hormonal or intrauterine methods, and both products have little effect on lipids or hemostasis. Introductory trials of MPArE2C in 12 000 women with 100 000 woman-months of experience confirmed the high efficacy of the product in routine use. The use of MPArE2C in a non-reusable injection device, Uniject W ŽBecton Dickinson, Franklin Lakes, NJ. is discussed. Once-a-month hormonal contraceptives have been shown to provide a safe contraceptive option for all women and an alternative for women who wish to use injectable formulations that cause less disruption in vaginal bleeding and minimal side effects. Q 1998 International Federation of Gynecology and Obstetrics Keywords: Contraception; Injectable; Once-a-month; Clinical trials

1. Introduction Some 40 years ago, dramatic changes occurred in the field of contraception with the advent of synthetic hormonal formulations. First came combined oral contraceptives ŽCOCs., followed not long afterwards by long-acting progestogens that U

Fax: q41 22 7914171.

were administered by intramuscular injection. These injectable products were made available in part to provide women with an option that would overcome problems of compliance with daily pill taking. However, from the outset, one of the major side effects associated with their use was disruption of vaginal bleeding patterns, an issue cited at a meeting convened by the World Health Organization ŽWHO. in 1981 as being the major reason for their discontinuation by users w1x. From

0020-7292r98r$19.00 Q 1998 International Federation of Gynecology and Obstetrics PII S0020-7292Ž98.00090-3

S44

P.E. Hall r International Journal of Gynecology & Obstetrics 62 Suppl. 1 (1998) S43]S56

the mid-1960s, investigators worked to overcome this problem by developing formulations containing both a progestogen and an estrogen w2]6x. Primarily because of the nature of available estrogen esters, combination progestogen]estrogen products had a duration of action of about 1 month. Since the first report w7x of a long-acting combined preparation containing 500 mg 17a-hydroxyprogesterone caproate plus 10 mg estradiol valerate, numerous preparations and dosages have been tested w2x. Of these, four preparations have been widely used. They are: v

v

v

v

17a-hydroxyprogesterone caproate plus estradiol valerate, an injectable which until the late 1980s was used extensively in China and neighboring countries, known as Chinese Injectable No. 1; 16 a ,17a-dihydroxyprogesterone acetophenide plus estradiol enanthate, produced by several companies in Latin America and marketed under various brand names; medroxyprogesterone acetate plus estradiol cypionate, which was originally developed as Cyclo-ProveraTM by Upjohn Company, Kalamazoo, Michigan, and subsequently as Cyclofem W following further development by WHO; and norethisterone enanthate plus estradiol valerate, developed by WHO and made available by Schering AG, Berlin, Germany as NETENrEV or NorigynonW , hereafter referred to as NET-ENrEV.

In 1994, WHO convened a meeting of experts who reviewed all available knowledge on once-amonth injectables but who focused principally on the latter two formulations w8x. The report of this meeting concluded that: ‘MPArE2C and NETENrEV are safe and effective products for fertility regulation which can be added to the existing range of contraceptive methods. They can be used by all potential contraceptive users providing precautions are taken to assess potential risk factors. They provide high efficacy, low incidence of side effects, and vaginal bleeding patterns better than those seen with progestogen-only injectables’. The full papers presented at the meeting

were published in two issues of the journal Contraception w1994;49Ž4,5.x. This paper will address these two recently available preparations but will focus in particular on 25 mg medroxyprogesterone acetate ŽMPA. plus 5 mg estradiol cypionate ŽEC. ŽCycloProveraTM , Pharmacia and Upjohn Company; MPArE2C W , Concept Foundation; HRP 112, WHO. hereafter referred to as MPArE2C. It will discuss the clinical and introductory studies supported by the UNDPrUNFPArWHOrWorld Bank Special Programme of Research, Development and Research Training in Human Reproduction, executed by WHO, Geneva, Switzerland, as part of its efforts to broaden contraceptive choices for women and men worldwide. The Special Programme began its activities on this topic in the late 1970s. Its goal was to develop a product that had been adequately studied for safety and efficacy, in contrast to the products then marketed in Latin America and China, and that could be made available to family planning programs around the world at an affordable price. Coutinho and de Souza w9x and subsequently others w2x had studied a formulation containing 25 mg of MPA and 5 mg of EC that was then developed further by Upjohn as Cyclo-ProveraTM . However, at the time of the WHO initiative, this product was not yet commercially available nor had any studies been undertaken with the other accessible long-acting progestogen, norethisterone enanthate ŽNET-EN.. Thus, after assessing the pharmacokinetics of three estradiol esters, the cypionate, the valerate, and the benzoate w10x, and the commercial availability of long-acting progestogens and the estradiol esters, the Special Programme began work on dose-finding studies of combinations of NET-EN plus estradiol valerate ŽEV. in comparison with MPA plus EC. At this time similar studies were also ongoing on NETEN plus EV by Sang Guo-wei in China, work which was subsequently incorporated into these development activities w11x. 2. Preclinical studies Toxicological data on depot MPA were thoroughly reviewed in 1981 by the Special Programme’s Toxicological Review Panel, which con-

P.E. Hall r International Journal of Gynecology & Obstetrics 62 Suppl. 1 (1998) S43]S56

sidered it safe for human use w1x. This panel had previously concluded that the beagle dog was an unsuitable model in which to observe potential adverse effects arising from the long-term use of progestogens in women, a view eventually supported by the United States Food and Drug Administration ŽFDA. w12x. Cookson w13x has reviewed the results on the toxicology of MPA and EC contained in MPArE2C, and Seibert and Gunzel w14x, those on the toxicology of NET-EN and EV contained in NET-ENrEV. Both products were considered safe for human use at the WHO meeting of experts in 1994 w8x. 3. Human studies A large number of clinical trials, including multicenter studies organized by the Special Programme, have been carried out in many countries with MPArE2C. These have included dose-finding and metabolic studies, Phase III clinical trials, and introductory studies. The total number of subjects participating in these studies and receiving MPArE2C and the total number of womanmonths of treatment are estimated to be:

Dose-finding and metabolic Phase III clinical trials Introductory studies Total

Number of subjects

Number of woman-months

Approx. 400

Approx. 2000

4534

44 160

) 12 200 ) 17 100

) 100 000 ) 146 000

Both products are given as deep intramuscular injections either into the deltoid or the gluteal muscle, once every 30 " 3 days Ž27]33 days.. MPArE2C is formulated as a microcrystalline suspension given in 0.5 ml of an aqueous solution, whereas NET-ENrEV is formulated in 1 ml of an oily solution of castor oil and benzyl benzoate, 60:40. 4. Pharmacological studies 4.1. Pharmacokineticr pharmacodynamic studies A series of studies was designed to investigate

S45

the absolute dosage and ratio of doses of the two components of the combination preparations containing MPA plus EC and NET-EN plus EV. The pharmacokinetic profiles of MPA in women receiving MPArE2C were studied by Fotherby et al. w15x. Wide intersubject variation in plasma steroid levels was seen, but ovulation was suppressed in all subjects. In the post-treatment phase, follicular activity returned in three of the 11 subjects by 28 days and in all subjects by 50 days. Plasma progesterone levels suggested that luteal activity did not occur until after day 63 following the last injection. The pharmacokinetics of MPA or NET and of estradiol ŽE 2 . in subjects receiving a single injection of either MPArE2C or NET-ENrEV were compared by Aedo et al. w16x. Although again large intersubject variation was observed, both preparations inhibited ovulation. Significant levels of both progestogens were measurable at 30 days. Levels of estradiol equivalent to a pre-ovulatory estradiol peak were noted with both esters, being higher with the valerate than the cypionate. In all cases estradiol levels returned to baseline before the end of the treatment cycle. Studies on the endometrium revealed suppressed or early proliferative patterns. Because of these studies, a reduction in dose of MPA to 12.5 mg Žhalf-dose. and EC to 2.5 mg was investigated in comparison with MPArE2C w17x. In the study, ovulation was assessed only during the third month of treatment and was inhibited by the reduced dose preparation in 19 out of 20 subjects. There were marked differences in the pharmacokinetic profiles between the centers participating in this study. A deterioration in bleeding patterns was observed with the half-dose preparation. A similar study was undertaken with NETENrEV and a half-dose preparation containing 25 mg of NET-EN plus 2.5 mg of EV w18x; although the latter preparation inhibited ovulation in all 23 subjects during the third treatment month, it also gave rise to a deterioration in bleeding patterns. A final dose-finding study was undertaken in which the dose of progestogen was halved Ž12.5 mg MPA or 25 mg NET-EN. and the estrogen

S46

P.E. Hall r International Journal of Gynecology & Obstetrics 62 Suppl. 1 (1998) S43]S56

maintained at 5 mg w18x. This change in dose resulted in breakthrough ovulation with both preparations during the third treatment month in 10 out of 24 subjects taking the MPA-containing preparation and in 7 out of 22 taking the NETEN-containing preparation. Thus, the pharmacokinetic studies led to the choice of formulations of MPArE2C and NETENrEV that caused suppression of ovulation and the least disturbance in vaginal bleeding patterns.

} ‘desire for pregnancy’ } were followed for a period of 12 months or until they had a confirmed pregnancy. The cumulative pregnancy rate was 81.9% after 12 months, similar to that observed with other methods w21x. A very low pregnancy rate was noted at the end of the first non-treatment month, although from this point it increased rapidly. No correlation was noted between the number of months to become pregnant and either duration of MPArE2C use or the weight of the subject.

4.2. Return to o¨ ulation and fertility 5. Phase III clinical trials With both MPArE2C and NET-ENrEV, pharmacodynamic studies showed follicular activity by 41]49 days and subsequent competent luteal activity by 59]87 days w16x. In a study of 21 women who received MPArE2C for 3 months, 52% ovulated during the first post-treatment month and 71% in the second post-treatment month w17x. In a similar study of 21 women who were taking NET-ENrEV, 19% ovulated in the first post-treatment month and 67% in the second w19x. After 2 years of use of either MPArE2C or NET-ENrEV, 54% of the NET-ENrEV recipients ovulated at least once during the 3 months after discontinuation, and 60% of the MPArE2C recipients ovulated by the third follow-up month w19x. Return of fertility was examined in the introductory studies, described later, that were conducted in Latin America w20x. A total of 70 subjects who had discontinued for the stated reason

Three major multicenter comparative clinical trials have been undertaken on MPArE2C and NET-ENrEV and supported by the Special Programme. The first was a 17-center, 12-country study w22x; the second was conducted in 11 clinics in Egypt Žas reported in w8x.; and the third was conducted in 15 clinics in three provinces in China w23x. A comparative study of depot-medroxyprogesterone acetate ŽDMPA. and MPArE2C was also undertaken in Vietnam w24x. Key results from these studies are shown in Table 1. 5.1. Contracepti¨ e efficacy MPArE2C and NET-ENrEV have both proved to be highly effective contraceptives: the 12-month pregnancy rates were 0.4% or less for NETENrEV and 0.2% or less for MPArE2C ŽTable 1.. There were no statistically significant differ-

Table 1 Summary of comparative Phase III studies of MPArE2C and NET-ENrEV: 12-month life-table rates Žper 100 women. of reasons for discontinuation Event

13 countries w22x

Egypt w8x

China w23x

Vietnama

MPArE2C NET-ENrEV MPArE2C NET-ENrEV MPArE2C NET-ENrEV MPArE2C Unintended pregnancy 0.0 Amenorrhea 2.1 Bleeding-related problems 6.3 Other medical reasons 6.2 Non-medical reasons 15.0 Total discontinuations 35.4 Number of women enrolled 1168 Woman-months of use 10 969 a

0.2 1.6 7.5 6.6 16.6 36.8 1152 10 608

Was part of a comparative study with DMPA w24x.

0.2 2.7 7.4 7.8 12.4 38.9 1111 10 492

0.4 1.4 11.5 4.7 12.8 38.0 1093 10 033

0.1 5.2 12.7 3.4 8.7 26.4 1955 19 765

0.3 0.8 7.9 3.6 8.1 18.8 1960 20 688

0 3.5 7.4 3.4 10.0 25.7 300 2934

P.E. Hall r International Journal of Gynecology & Obstetrics 62 Suppl. 1 (1998) S43]S56

ences between the rates found with the two preparations, although rates were slightly higher for NET-ENrEV. There was one ectopic pregnancy with NET-ENrEV. In total, there were five pregnancies in 44 160 woman-months of exposure with MPArE2C and 12 pregnancies in 41 329 woman-months of exposure with NET-ENrEV in these studies. Many of the pregnancies occurred in the first month of treatment, although the timing of subsequent injections Žonce every 30 " 3 days. is likely to be extremely important to maintaining high contraceptive efficacy. 5.2. Reasons for discontinuation Twelve-month life-table discontinuation rates for amenorrhea and bleeding-related problems with NET-ENrEV and MPArE2C in Phase III studies are summarized in Table 1. Although variation from the normal, regular pattern of vaginal bleeding accounted for the majority of women discontinuing once-a-month injectable use for medical reasons, the rates of discontinuation due to those variations were much lower than those usually reported with long-acting progestogenonly injectables w25,26x. In these studies, rates for discontinuation because of amenorrhea in these studies varied between 0.8% and 1.6% for NETENrEV and 2.1% and 5.2% for MPArE2C, compared with discontinuation for bleeding-related reasons of 7.5% and 11.5% for NETENrEV and 6.3% and 12.7% for MPArE2C. The main bleeding disturbances reported were heavy, prolonged, and irregular bleeding. Major differences have been noted in the discontinuation rates for different stated reasons in individual centers in most countries. For example, in the Phase III studies, rates of discontinuation due to amenorrhea varied in different centers between 0 and 17.5%, whereas rates of discontinuation due to bleeding disturbances varied between 0 and 34.0%. From an analysis of menstrual diary cards Žsee below., it is obvious that only part of this variation can be explained by actual differences in menstrual patterns; the most significant influence is probably cultural and counseling factors w22x. The discontinuation rates for other Žnon-

S47

menstrual . medical reasons were low and were similar in all three studies, as shown in Table 1. The reasons given were similar for both formulations and included headaches, dizziness, body aches, mastalgia, and body weight changes. The reasons appeared to vary with ethnic background. Body weight tended to increase but was less than a mean of 1 kg after 12 months in most centers. Non-medical reasons for discontinuation included study termination, moving away, planned pregnancy, and no further need; discontinuation rates varied from 8.1% to 16.6% in the three studies. The total 12-month discontinuation rates, which ranged from 18.8% to 38.9%, were significantly lower than the rates in similar studies of the long-acting progestogen-only injectables, DMPA and NET-EN w25,26x. 5.3. Vaginal bleeding patterns In the Phase III trials supported by the Special Programme, women were requested to use daily menstrual diary cards to record the occurrence of bleeding and spotting. Analysis of a large data set collected from ‘normal women’ allowed the Special Programme to generate a precise statistical evaluation of the variations in normal vaginal bleeding patterns w27x. This study also enabled the following definitions of vaginal bleeding patterns using a 90-day reference period to be constructed: Amenorrhea: Infrequent bleeding: Frequent bleeding: Irregular bleeding: Prolonged bleeding:

No bleeding Fewer than two bleedingr spotting episodes More than four bleedingr spotting episodes A range of lengths of bleeding-free intervals exceeding 17 days At least one bleedingrspotting episode lasting 10 days or more

These definitions were used to show that over 1 year there was little difference with respect to the incidence of amenorrhea, infrequent, frequent, irregular and prolonged bleeding ŽTable 2. between women using MPArE2C and those using NET-ENrEV in the WHO multicenter study w22x. Women who had used either contraceptive

P.E. Hall r International Journal of Gynecology & Obstetrics 62 Suppl. 1 (1998) S43]S56

S48

Table 2 Percentage Ž%. of women experiencing different types of bleeding patterns Group

Days

Number of diaries

Untreated

1]90 91]180 181]270 271]360

3893 3893 3893 3893

1.3 1.5 1.3 1.6

3.4 2.9 2.8 3.1

0.2 0.3 0.1 0.3

4.5 4.8 5.4 8.6

MPArE2C

1]90 91]180 181]270 271]360

1001 885 802 730

0.1 0.2 1.1 2.3

0.1 3.4 5.4 3.7

22.3 3.3 2.8 6.5

NET-ENrEV

1]90 91]180 181]270 271]360

1000 860 766 713

0.2 0.6 1.3 2.0

0.1 2.2 2.9 5.0

DMPA

1]90 91]180 181]270 271]360

509 406 311 241

10.6 23.9 37.0 38.6

15.7 25.8 24.8 27.8

a

Amenorrhea

Infrequent bleeding

Frequent bleeding

Irregular bleeding

Prolonged bleeding

Total women showing variations from regular patterna

Women showing regular patterns

2.6 2.3 2.6 4.3

9.7 9.2 9.9 14.9

90.3 90.8 90.1 85.1

39.6 23.5 25.4 13.6

20.8 13.3 9.4 10.1

57.1 36.8 38.7 30.0

43.0 63.2 61.3 70.0

29.6 5.5 4.9 6.2

34.6 25.2 24.8 14.6

16.2 11.1 12.6 12.7

52.8 37.2 36.7 31.6

47.2 62.8 63.3 68.4

17.7 10.5 8.3 6.6

46.0 35.7 27.7 17.9

43.4 27.7 17.3 16.5

91.0 93.1 93.6 91.7

9.0 6.9 6.4 8.3

Some subjects may have appeared in more than one category of bleeding pattern.

had a very low incidence of amenorrhea, which increased only slightly during 1 year of use, quite different from the dramatic increase in amenorrhea in DMPA users who participated in a previous WHO study w24x. This analysis shows that both MPArE2C and NET-ENrEV users have a high incidence of frequent, irregular, and prolonged bleeding in the first 3 months of use; however, this high incidence is due to the additional bleeding episode associated with the first injection. The incidence of these variations decreased greatly during 1 year of use. By the end of 1 year, some 70% of MPArE2C and NET-ENrEV users experienced regular bleeding patterns compared with only 8% of DMPA users. The incidence of most variations from regular patterns tended to decrease with time in once-a-month injectable users. Compared with women who did not use contraceptives, the longer-term once-a-month injectable users had a lower incidence of amenorrhea, an equal incidence of infrequent bleeding, and a significantly

increased incidence of frequent, irregular, and prolonged bleeding. A similar analysis was undertaken of data from the Chinese study w28x. The number of bleedingrspotting days over four periods was similar for the two groups, but there were more spotting days and there was greater individual variability among MPArE2C users. The irregular bleeding in the first 3 months was 21.1% for NET-ENrEV and 26.1% for MPArE2C, decreasing to 4.4% and 7.7%, respectively, in the last 3 months of the treatment year. In Egypt, the opposite was observed; fewer bleeding problems were observed with MPArE2C than with NETENrEV w8x. 6. Metabolic studies 6.1. Coagulation and fibrinolysis The Special Programme undertook a large multicenter study comparing a COC Ž35 m g ethinyl

P.E. Hall r International Journal of Gynecology & Obstetrics 62 Suppl. 1 (1998) S43]S56

estradiol combined with 1 mg norethindrone. with MPArE2C and NET-ENrEV and investigating 11 parameters of hemostasis w29x. The COC induced increases in fibrinogen factor VII and X activities and plasminogen. These prothrombotic changes were reflected in a shortening of the activated partial thromboplastin time. By contrast, protein C was increased by nearly 10%. Neither of the injectable contraceptives induced a rise in procoagulant factors. Both induced a minor reduction in factor VII and X activities, particularly NET-ENrEV, and this was thought to reflect the more anti-estrogenic effect of NET compared with that of MPA. An increase in tissue plasminogen activator inhibitor was noted in both groups, reaching 12% by the end of the treatment period. Small decreases in antithrombin III activity and in protein C were not considered clinically relevant. Overall, the changes noted with both once-amonth formulations were exceedingly small, with NET-ENrEV having less effect than the COC on hemostasis and MPArE2C having even less. The once-a-month injectables did not induce an increase in procoagulants and had a negligible effect on the fibrinolytic system. 6.2. Lipid metabolism One small study found no change in total cholesterol but a reduction in HDL-cholesterol ŽHDL-C. after 6 months of MPArE2C use; no significant changes were found with NET-ENrEV w30x. However, to fully assess the effects of MPArE2C and NET-ENrEV on lipids and apolipoproteins, the Special Programme undertook a large comparative study in Hangzhou, Havana, Jakarta and Shanghai. The study investigated blood lipid levels in the pretreatment cycle and third and ninth treatment cycles w31x. With both MPArE2C and NET-ENrEV, total cholesterol decreased during the period of treatment, particularly during the first week after each injection. The effect of the estrogen dominated during the first week of the injection interval, when levels of LDL-cholesterol ŽLDL-C. and apolipoprotein B decreased by up to 10%, to return to initial levels at the end of each injection

S49

interval. In the second and third weeks of the injection interval, decreases in HDL-C and apolipoprotein A-I were observed, reflecting a dominant progestogen effect. Triglyceride levels were decreased, particularly 1 week after each injection. All parameters had returned to pretreatment levels when assessed 3 months after discontinuation of the contraceptive. The effect of MPArE2C on all these parameters was less than that of NET-ENrEV. With both preparations, the HDL-Crtotal cholesterol ratio was increased in the first week after each injection, to decrease by no more than 6% from baseline in the third week. Conversely, the LDL-Crtotal cholesterol ratio was decreased in the week that followed each injection by 4]6% from baseline in the third week. The HDL-CrLDL-C ratio increased during the first week after each injection and decreased in the third week. The apolipoprotein AIrB and AIrAII ratio varied in parallel with the HDL-Crtotal cholesterol ratio and did not vary by more than 10% during the period of follow-up. All ratios returned to baseline by the third post-treatment month. The minor changes observed in lipid and apolipoprotein levels with MPArE2C and NETENrEV are not clinically significant. The changes closely followed the pharmacokinetic profiles of the progestogen and estrogen components of the formulations and their relative ratios; for example, estradiol returned to basal levels in the second half of each treatment interval. These minor changes had returned to baseline levels when assessed 3 months after discontinuation of treatment. Although no studies of a comparable size have been undertaken with progestogen-only injectable contraceptives, the data suggest that the once-a-month combination products have considerably less effect on lipid metabolism, particularly on HDL-C w31x. 6.3. Introductory studies Introductory trials of MPArE2C were undertaken to determine whether, under routine service conditions, the use-effectiveness and reasons for discontinuation were similar to those noted in earlier clinical studies of safety and efficacy Žde-

S50

P.E. Hall r International Journal of Gynecology & Obstetrics 62 Suppl. 1 (1998) S43]S56

scribed previously .. Introductory trials of MPArE2C were undertaken in Brazil, Chile, Colombia, Indonesia, Jamaica, Mexico, Peru, Thailand and Tunisia. More than 12 000 women were studied and more than 100 000 womanmonths of experience accumulated w32,33x. The cumulative 12-month life-table discontinuation rates for all reasons are shown in Table 3 and ranged from 33.5% to 75.6%. The rates of discontinuation in all countries were similar or lower than those found for DMPA in general use, except in Thailand, where discontinuation rates were higher than for general use of hormonal contraceptives. With regard to efficacy, in Mexico more than 20 000 woman-months of experience were accumulated with only one pregnancy attributed to the method, whereas in Indonesia, four pregnancies occurred. In Brazil, Chile, Colombia, Jamaica, Peru, Thailand and Tunisia, no pregnancies were reported. Thus, in these studies there were only five pregnancies in a total experience of more than 100 000 woman-months of use. Reasons for discontinuation varied greatly between countries ŽTable 3.. The Phase III clinical trial of MPArE2C had shown a major difference between centers in discontinuation rates for

bleeding-related reasons and amenorrhea. Similar differences were reflected in the introductory trials, in which a large range of discontinuation rates was observed for bleeding-related reasons, from 1.4% in Indonesia to 24.8% in Tunisia, and from 1.4% to 13.2% in the same countries for amenorrhea. Lack of tolerance even to minor spotting episodes was observed in Tunisia, where bleeding-related reasons were the most frequent reason for discontinuation. In Thailand, women were asked whether they had experienced any bleeding since the last injection. For those women who had switched from DMPA because of amenorrhea, some 70% experienced a menstrual bleeding episode by the end of the third month of treatment with MPArE2C. Discontinuations for other medical reasons reflected the major concerns about side effects. The rates were lowest in Mexico and Indonesia and highest in Chile, Colombia, Peru and Tunisia. In Indonesia, the most common reason was dizziness. In Thailand, weight gain, headaches and dizziness were the most common reasons. In the South American countries, headaches and weight gain were the primary reasons. In Tunisia, a large number of women discontinued because of

Table 3 MPArE2C introductory studies: number of subjects discontinuing and cumulative 12-month life-table discontinuation rates per 100 women by country and major reasons Reasons for discontinuation

Indonesia

Jamaica

Thailand

Tunisia

Mexico

Brazil

Chile

Colombia

Peru

No. Rate

No. Rate

No. Rate

No. Rate

No.

Rate

Rate

Rate

Rate

Rate

All reasons All medical reasons Pregnancy Bleeding-related Amenorrhea Other medical reasons All personal reasons Inconvenience and timing Moved away Desire for pregnancy No further need Negative perceptions Change of method Other personal reasons Lost to follow-up

376 82 4 13 11 49 222 67 43 65 6 ] 1 40 70

150 42 0 15 3 23 75 30 9 6 9 ] 14 7 33

762 231 0 49 43 139 430 153 123 66 45 ] 33 10 101

621 353 0 144 63 146 198 51 36 27 ] 60 ] 24 70

2671 ] 1 ] ] ] ] ] ] ] ] ] ] ] ]

75.6 22.8 0.1 13.7 5.9 4.8 69.1

49.2 ] 0 3.4 9.2 7.8 42.2

48.0 ] 0 6.7 5.1 19.7 29.2

57.7 ] 0 8.1 7.7 26.3 35.8

49.3 ] 0 6.4 8.9 20.5 26.3

U

U

U

U

U

11.9

5.5

8.5

7.6

3.9

U

U

U

U

U

U

U

U

U

U

U

U

U

4.7

10.2

2.3

24.3

33.5 8.5 0.5 1.4 1.4 5.0 21.6 7.2 4.7 7.2 0.6 ] 0.1 3.8 6.9

40.4 14.8 0 5.8 1.2 7.7 24.0 8.9 4.0 2.5 4.4 ] 4.2 2.6 7.8

58.3 22.8 0 5.1 5.0 14.3 39.5 16.2 13.3 7.7 4.6 ] 4.1 1.2 10.5

Note. An asterisk indicates included under ‘other personal reasons’.

71.8 52.3 0 24.8 13.2 26.5 31.5 8.1 6.5 6.2 ] 11.2 ] 4.1 11.9

U

U

29.6 49.6 7.0

U

U

U

U

U

U

U

U

P.E. Hall r International Journal of Gynecology & Obstetrics 62 Suppl. 1 (1998) S43]S56

headaches, backache, nervousness, heaviness of limbs and weight gain. Personal reasons for discontinuation included those not due directly to method-attributable side effects. A woman’s decision to discontinue for personal reasons may be influenced by other users or members of the community, may reflect the treatment or perceived treatment she has received from service delivery staff, or may be due to the inconvenience of the services. Personal reasons were cited more often than in the Phase III clinical trials. They included ‘inconvenience and timing’, ‘moved away’, ‘desire for pregnancy’, and in the case of Thailand, ‘no further need’ and ‘change of method’. In Tunisia, the reasons included ‘negative perceptions’ on the part of both users and providers. ‘Inconvenience and timing’ reflected problems caused by the injection being available only in a limited number of clinics, the distance to those clinics, their opening hours, the need to visit once a month, and mobility of the populations. This last reason contributed significantly to discontinuations in South America, whereas discontinuation for this reason ranged from 7.2% in Indonesia to 16.2% in Thailand. In several countries a large number of women migrated to the cities or to other provinces, often for limited periods of time. ‘Moved away’ was an important reason in Thailand Ž13.3%.. Although ‘desire for pregnancy’ can be an easy reason to give for discontinuation, it also indicates that the method was used for spacing purposes, because this reason was cited more frequently with time. ‘No further need’ reflected divorce, death of husband, husband having undergone a vasectomy, or temporary separation of partners. The studies confirmed the high efficacy of MPArE2C but showed, as did the Phase III clinical trials, major variations between countries with respect to discontinuation for bleeding-related reasons, amenorrhea, and other medical reasons. This observation underscores the importance of counseling women regarding menstrual changes and other side effects prior to and during use of monthly injectables. Other reasons for discontinuation were influenced by the limited

S51

availability of the method and other service delivery issues. 7. Medical eligibility criteria To assist countries in updating their medical guidelines on family planning, WHO convened two meetings of scientific experts in March 1994 and May 1995. The experts reviewed all data that had become available over the last 10 years from clinical and epidemiological research on contraceptive methods. The participants also recommended medical eligibility criteria for different contraceptive methods, criteria that would protect men and women from the potential adverse effects of contraceptives by an adequate margin of safety while, at the same time, not unduly denying them a choice of suitable methods w34x. The document defined conditions that affected the eligibility for use of each family planning method and classified them under the following categories: Category 1

Category 2

Category 3

Category 4

Conditions for which there is no restriction in the use of the family planning method. Conditions for which the advantages of using the method generally outweigh the theoretical or proven risks. Conditions for which the theoretical or proven risks usually outweigh the advantages of using the method. Conditions for which the use of the method represents an unacceptable health risk.

Categories 1 and 4 are self-explanatory. For persons with conditions classified under Category 2, the method can be used but careful follow-up may be required and the condition should be a consideration during counseling on method choice. For persons with conditions classified under Category 3, careful clinical judgment is required, taking into account the severity of the condition and the availability, practicality, and acceptability of alternative methods. If the method is used, it should be the method of last choice and careful, routine follow-up care must be provided. The criteria as applied to the use of once-a-month injectable contraceptives are shown in Tables 4]7. The criteria listed in Tables 4]7 are conserva-

S52

P.E. Hall r International Journal of Gynecology & Obstetrics 62 Suppl. 1 (1998) S43]S56

Table 4 Medical eligibility criteria: WHO Category 1 } no restrictions on use

Table 5 Medical eligibility criteria: WHO Category 2 } general use with caution

Age: menarche to 40 years Nulliparous Parous Obesity Gynecologicrobstetric conditions History of pre-eclampsia History of ectopic pregnancya Post-abortion G 21 days postpartum Žnon-breastfeeding . Irregular menstrual pattern with or without heavy bleeding Severe dysmenorrheaa Endometriosis Pelvic inflammatory disease Žhistory or current.a Benign breast diseasea Family history of breast cancer Benign ovarian tumors Cervical ectropionrerosion Uterine fibroids Endometrial or ovarian cancera Gestational trophoblastic disease Žbenign or malignant. History of pregnancy-related diabetes Chronic diseaserother conditions Thyroid disease Epilepsyb Superficial varicose veins Mild headaches Viral hepatitis Žcarrier; not active case. Schistosomiasis Fibrosis of the liver Malaria Iron-deficiency anemiaa Tuberculosis } non-pelvic or pelvicc Prior pelvic surgery STDrHIV risk Žadvise condom use. STDs Žhistory or current . HIV-positive or AIDS

Age ) 40 years Smoker - 35 years Severe headaches Žincluding migraine without focal neurological symptoms. Gynecologicrobstetric conditions Breast-feeding ŽG 6 months postpartum. Breast disease } undiagnosed mass History of pregnancy-related cholestasis Cervical cancer or precancerous cervical lesions Cardiovascular conditions Known hyperlipidemias Ždepending on type and severity. Mild hypertension Žblood pressure - 160r100. if blood pressure can be monitored periodically, if not, this is Category 3 Superficial thrombophlebitis Uncomplicated valvular heart disease Chronic diseasesrother conditions Thalassemia Sickle cell disease Insulin and non-insulin-dependent diabetes Žuncomplicated. Surgically treated or asymptomatic biliary tract disease Current or medically treated biliary tract disease; Category 4 if severe Mild cirrhosis Žcompensated.

a

Combined COCs provide protection against these conditions. If not using certain anticonvulsant drugs on a long-term basis. Certain antiseizure drugs Žphenytoin, carbamezapine, barbiturates, and primadone. are likely to reduce the efficiency of hormonal contraceptives. c If not using certain antibiotics. Certain antibiotic drugs Žrifampicin and griseofulvin. are likely to reduce the efficiency of hormonal contraceptives. b

tive and, in some instances, because there are no long-term epidemiological data on combined progestogenrestrogen formulations, they reflect the criteria for COCs. For example, extensive studies have been done on the effects of progestogen-only contraception on breast milk,

and no ill effects were noted in the infants of breast-feeding mothers, whereas COCs decrease the quality and quantity of breast milk. There are, however, no data on combined injectable formulations. Combined progestogen]estrogen injectables can be started after weaning, but further studies are needed to determine whether these products can be safely used during breast-feeding. Thus, WHO recommends that in the absence of evidence, they be treated similarly to COCs. As for once-a-month injectable contraceptives, they can be started within the first 5 days after menstruation begins and no backup method is required. After that time, a woman can begin to use the contraceptive if she has not been exposed to the risk of pregnancy and if she uses a backup method for the next 7 days. Women who have amenorrhea but in whom pregnancy has been excluded can start immediately. Subsequent injections must be given every 30 " 3 days Ž27]33 days..

P.E. Hall r International Journal of Gynecology & Obstetrics 62 Suppl. 1 (1998) S43]S56 Table 6 Medical eligibility criteria: WHO Category 3 } method of last choice Light smoker, age ) 35 years Gynecologicrobstetric conditions Breast-feeding Ž6 weeks]6 months postpartum. - 21 days postpartum Žnon-breast-feeding . History of breast cancer and no evidence of disease in past 5 years Unexplained vaginal bleeding Cardiovascular conditions History of hypertension; current blood pressure not known Hypertension Žblood pressure 160]179r100]109.a Known hyperlipidemiasb Chronic diseasesrother conditions History of combined OC-related cholestasis Use of certain antibiotic or antiseizure medications Nephropathyrretinopathyrneuropathy Žif condition very severe, this is Category 4. Active viral hepatitis; Category 4, if severe Liver tumors Žbenign or malignant.; Category 4, if severe malignancy Severe cirrhosis Ždecompensated. a

If blood pressure cannot be monitored periodically, this is Category 4. b Could be Category 2, depending on the type and severity of the condition.

8. Use of MPAr r E2C in Uniject, a non-reusable injection device Because of concerns about the use of syringes and needles particularly their reuse in countries where infectious diseases such as hepatitis B and the human immunodeficiency virus are prevalent, the Special Programme has been investigating non-reusable injection systems and, in particular, the Uniject device, which was recently licensed by the Program for Appropriate Technology in Health ŽPATH. to the Becton Dickinson Company. A study was undertaken to assess the effects of administering MPArE2C using prefilled Uniject devices w35x. A total of 480 injections were administered, equally divided between standard syringes and Uniject. The majority of women Ž82%. reported that they felt anxious before receiving the injection with Uniject. After injection, more than 80% reported having no anxiety about Uniject or about the injection technique. Some 96% of the participants reported having had slight to no pain

S53

Table 7 Medical eligibility criteria: WHO Category 4 } do not use Heavy smoker Ž) 20 cigarettesrday., age G 35 years Migraine headaches with focal neurological symptoms Gynecologicrobstetric conditions Known or suspected pregnancy Breast-feeding Ž- 6 weeks postpartum. Breast cancer Žcurrent . Cardiovascular conditions Moderatersevere hypertension Žblood pressure ) 180r110. Current or history of thromboembolic disorders or stroke Major surgery with prolonged immobilization Current or history of ischemic or complicated valvular heart disease Hypertension with vascular disease Chronic diseasesrother conditions Diabetes with certain vascular complications andror ) 20 years’ durationa Nephropathyrretinopathyrneuropathya a

Could be Category 3, depending on the type and severity of the condition.

with Uniject, although 5.6% expressed having had more pain with Uniject than with the standard syringe. The majority of service providers Ž90%. reported that Uniject was easy to activate and inject and that it was reassuring for users to know that the syringe and needle had never been used previously. Only two service providers had difficulties in activating the device. No other problems occurred during activation of the device or the injection procedure, and in all cases there was complete delivery of the drug from the Uniject device w35x. A second study was undertaken by the Brazilian group with MPArE2C in Uniject devices to evaluate women’s acceptance of and ability to self-administer the injectable contraceptive w36x. Of 88 women who agreed to participate in the study, 32 Ž31.4%. refused to administer the injections after training with oranges. Only 56 women Ž55% . ultimately injected themselves with MPArE2C, undertaking a total of 116 injections, all administered on the ventral side of the thigh. When nurses evaluated the women’s ability to activate the devices, they found that almost 90% were successful. Of those women, 93% were able to correctly self-administer the contraceptive from the Uniject device. Almost 60% of the women who self-injected

S54

P.E. Hall r International Journal of Gynecology & Obstetrics 62 Suppl. 1 (1998) S43]S56

said that they wished to continue administering the drug themselves; one third said that they were afraid. Some 12.5% of the women said that injection in the thigh was more painful than in the buttocks or arm. It would therefore appear that women can be trained to self-administer MPArE2C successfully with a non-reusable device such as Uniject w36x. In an age in which it is critical to use the best possible injection practices in service delivery systems, Uniject has been shown to be an appropriate non-reusable injection device for the administration of MPArE2C and may be equally applicable for other injectable contraceptives. 9. Availability of once-a-month injectable contraceptives Mesigyna W rNorigynon is manufactured in Mexico and is distributed by Schering AG, Berlin, Germany. It is registered in 36 countries Žof which 15 are Caribbean states . and is available principally in Latin America and a few Asian countries. Cyclofem W was licensed to the Concept Foundation, Bangkok, Thailand. It is currently manufactured under license by Aplicaciones Farmaceuticas, SA, in Mexico and P.T. Tunggal in Indonesia and will shortly be manufactured by a second Indonesian company and a company in Thailand. It is presently registered and distributed in 18 countries, primarily in Latin America and a few countries in Asia. It has also been licensed to Pharmacia and Upjohn Company, and a registration application has been filed with the US FDA. 10. Conclusions Injectable contraception is used as a method of contraception by more than 12 million women worldwide. The majority is using highly effective progestogen-only preparations containing either DMPA or NET-EN. The main reasons for discontinuation of these products relate to significant disruption of vaginal bleeding. To ameliorate these problems, once-a-month injectables combining a progestogen with an estrogen have been developed. In recent times, two new preparations have become available through research sup-

ported by the UNDPrUNFPArWorld Bank Special Programme of Research, Development and Research Training in Human Reproduction. These are CyclofemrCyclo-Provera Ž25 mg MPA plus 5 mg EC. and MesigynarNorigynon Ž50 mg NET-EN plus 5 mg EV.. A series of pharmacokineticrpharmacodynamic studies were conducted to explore dosage and ratios of MPA and NET-EN and of the respective estrogens. These studies determined the final doses of the constituents of MPArE2C and NET-ENrEV and were chosen to give high efficacy and maintain acceptable vaginal bleeding patterns. Several large-scale clinical Phase III studies have confirmed the high efficacy of both MPArE2C and NET-ENrEV. The pregnancy life-table rates were all below 0.4% for NETENrEV and below 0.2% for MPArE2C. More than two thirds of women had predictable regular cycles. The discontinuation rates for bleeding-related problems were less than half of those noted for progestogen-only injectables, and the discontinuation rates for amenorrhea were low. The time to return to fertility was similar to that for other commonly used methods and more rapid than that for the progestogen-only injectable contraceptives. No adverse or clinically relevant metabolic changes have been observed in studies of MPArE2C and of NET-ENrEV. The minor changes observed in lipid and apolipoprotein levels up to 9 months of use of these products closely followed the pharmacokinetic profiles of the progestogen and estrogen components of the formulations. These changes appear to be considerably less than with progestogen-only injectable contraceptives. With regard to effects on coagulation or fibrinolytic factors, the changes noted with both once-a-month formulations were extremely small, NET-ENrEV having less effect than a COC on hemostasis and MPArE2C having even less. The once-a-month injectables did not induce an increase in procoagulants and had a negligible effect on the fibrinolytic system. Introductory trials of MPArE2C in nine countries studied more than 12 000 women with more than 100 000 woman-months of experience. These

P.E. Hall r International Journal of Gynecology & Obstetrics 62 Suppl. 1 (1998) S43]S56

studies confirmed the high efficacy of the product in family planning services and addressed management and service issues. For MPArE2C, some 17 100 women have participated in, and contributed more than 146 000 woman-months of experience to, dose-finding and metabolic studies, Phase III clinical trials, and introductory studies. Although these results suggest that MPArE2C and NET-ENrEV can be used by most women, medical eligibility criteria are given that are rather more conservative, reflecting those for COCs since there are no long-term epidemiological data yet available on combined progestogenrestrogen formulations. In an age in which it is critical to use the best possible injection practices in service delivery systems, the non-reusable injection device Uniject has been shown to be an appropriate delivery device for the administration of MPArE2C. Once-a-month hormonal contraceptives have been shown to provide many women with a safe and highly effective contraceptive option with minimal side effects as well as an option for those who wish to use injectable formulations but cannot tolerate the disruption of vaginal bleeding experienced with the widely used progestogenonly formulations. Acknowledgements The author should like to thank, in particular, Dr Catherine d’Arcangues of the Special Programme of Research, Development and Research Training in Human Reproduction, who was responsible for many of the studies discussed and for her support and assistance in these activities. Thanks are also due to the 17 000 women who volunteered to participate in the studies on MPArE2C, particularly those who had innumerable venipunctures for the pharmacological and metabolic studies; to the investigators, doctors, and nurses who looked after them; and to the UNDPrUNFPArWHOrWorld Bank Special Programme of Research, Development and Research Training in Human Reproduction for funding the studies.

S55

References w1x World Health Organization. Facts about injectable contraceptives: memorandum from a WHO meeting. Bull WHO 1982;60:210. w2x Benagiano G. Long acting systemic contraceptives in regulation of human fertility. In: Diczfalusy E, editor. Regulation of human fertility. Copenhagen: Scriptor, 1977:323]360. w3x Toppozada M. The clinical use of monthly injectable contraceptive preparations. Obstet Gynecol Surv 1977;32:335]347. w4x Hall PE, Fraser IS. Monthly injectable contraceptives. In: Mishell DR, editor. Long acting steroid contraception. New York: Raven Press, 1983;65]88. w5x Hall PE. Once-a-month injectable contraceptives. IPPF Med Bull 1987;21:1]2. w6x Liskin L, Blackburn R. Hormonal contraception: new long acting methods. The Johns Hopkins University.. Popul Rep 1987;15:K57]K87. w7x Siegel I. Conception control by long-acting progestogens. Preliminary report. Obstet Gynecol 1963; 21:666]668. w8x World Health Organization. Facts about once-a-month injectable contraceptives: memorandum from a WHO meeting. Bull WHO 1994;70:677]689. w9x Coutinho EM, de Souza JC. Conception control by monthly injections of medroxyprogesterone suspension and a long-acting oestrogen. J Reprod Fertil 1968;15:209]214. w10x Oriowo MA, Landgren B-M, Stenstrom ¨ B, Diczfalusy E. A comparison of the pharmacokinetic properties of three estradiol esters. Contraception 1980;21:415]424. w11x Sang G-W. Pharmacodynamic effects of once-a-month combined injectable contraceptives. Contraception 1994;49:361]385. w12x Jordan A. FDA requirements for non-clinical testing of contraceptive steroids. Contraception 1992;46:499]509. w13x Cookson KM. The parenteral toxicity of MPArE2C. Contraception 1994;49:335]345. w14x Seibert B, Gunzel P. Animal toxicity studies performed ¨ for risk assessment of the once-a-month injectable cont r a c e p t iv e N E T - E N r E V . C o n t r a c e p t io n 1994;49:303]333. w15x Fotherby K, Benagiano G, Toppozada HK, Abdel-Rahman A, Navaroli F, Arce B et al. A preliminary pharmacological trial of the monthly injectable contraceptive Cycloprovera. Contraception 1982;25:261]272. w16x Aedo A-R, Landgren B-M, Johannisson E, Diczfalusy E. Pharmacokinetic and pharmacodynamic investigations with monthly injectable contraceptive preparations. Contraception 1985;31:453]469. w17x World Health Organization. Task force on long-acting systemic agents for fertility regulation, special programme of research, development and research training in human reproduction. A multicentred pharmacokinetic, pharmacodynamic study of once-a-month in-

S56

w18x w19x w20x

w21x w22x

w23x

w24x

w25x

w26x

w27x

P.E. Hall r International Journal of Gynecology & Obstetrics 62 Suppl. 1 (1998) S43]S56 jectable contraceptives: I. Different doses of HRP112 and of Depoprovera. Contraception 1987;36:441]457. Garza-Flores J. Pharmacokinetics of once-a-month injectable contraceptives. Contraception 1994;49:347]359. Bassol S, Garza-Flores J. Return of ovulation upon discontinuation of once-a-month injectable contraceptives. Contraception 1994;49:441]453. Bahamondes L, Lavin P, Ojeda G, Petta C, Diaz J, Maradiegue E et al. Return of fertility after discontinuation of the once-a-month injectable contraceptive cyclofem. Contraception 1997;55:307]310. Zinaman MJ, Clegg ED, Brown CC, O’Conner J, Selevan SG. Estimates of human fertility and pregnancy loss. Fertil Steril 1996;65:503]509. World Health Organization. Task force on long-acting systemic agents for fertility regulation. A multicentred phase III comparative study of two hormonal contraceptive preparations given once-a-month by intramuscular injection: I. Contraceptive efficacy and side-effects. Contraception 1988;37:1]20. Sang GW, Shao QX, Ge RS, Ge JL, Chen JK, Song S et al. A multicentred phase III comparative clinical trial of NET-ENrEV, MPArE2C and Injectable No. 1 given monthly by intramuscular injection to Chinese women. I. Contraceptive efficacy and side effects. Contraception 1995;51:167]183. Cuong DT, My Hyong NT. Comparative phase III clinical trial of two injectable contraceptive preparations, depot medroxyprogesterone acetate and MPArE2C, in Vietnamese women. Contraception 1996;54:169]179. World Health Organization. Task force on long-acting systemic agents for fertility regulation. Multinational comparative clinical trial of long-acting injectable contraceptives: norethisterone enanthate given in two dosage regimens and depot medroxyprogesterone acetate. Contraception 1983;28:1]20. World Health Organization. Task force on long-acting systemic agents for fertility regulation. A multicentred phase III comparative clinical trial of depot medroxyprogesterone acetate given three-monthly at doses of 100 mg or 150 m: I. contraceptive efficacy and side-effects. Contraception 1986;34:223]235. Belsey EM, Pinol APY. Task force on long-acting systemic agents for fertility regulation. Menstrual patterns in untreated women. Contraception 1997;55:57]65.

w28x Sang GW, Shao QX, Ge RS, Ge JL, Chen M, Song S et al. A multicentred phase III comparative clinical trial of NET-ENrEV, cyclofem and injectable No. 1 given monthly by intramuscular injection to Chinese women. II. The comparison of bleeding patterns. Contraception 1995;51:185]192. w29x World Health Organization, Task force on long-acting systemic agents for fertility regulation. A comparative study of the effects of two once-a-month injectable steroidal contraceptives and a combined oral contraceptive on haemostasis. Contraception 1998 Žin press.. w30x Haiba NA, El Habashy MA, Said SA, Darwish EA, Abdel-Sayed WS, Nayel SE. Clinical evaluation of two monthly injectable contraceptives and their effects on some metabolic parameters. Contraception 1989; 39:619]632. w31x World Health Organization, Task force on long-acting systemic agents for fertility regulation. A comparative study of the effects of two once-a-month injectable steroidal contraceptives Ž NE T-E NrE V W and MPArE2C W . on lipid and lipoprotein metabolism. Contraception 1997;56:193]207. w32x Hall PE. WHO. The introduction of MPArE2C in National family planning programmes: experience from studies in Indonesia, Jamaica, Mexico, Thailand and Tunisia. Contraception 1994;49:489]507. w33x Hall PE, Bahamondes L, Diaz J, Petta C, Marchi NM, Nakagava H et al. Introductory studies of the once-amonth injectable contraceptive MPArE2C in Brazil, C h ile , C olom b ia a n d P e ru . C on tra ce ption 1997;56:353]359. w34x World Health Organization. Improving quality of care in reproductive health: medical eligibility criteria for family planning methods. Geneva: WHO, 1996. w35x Bahamondes L, Marchi NM, Cristofoletti ML, Nakagava HM, Pellini E, Araujo F et al. Uniject as a delivery system for the once-a-month injectable contraceptive cyclofem in Brazil. Contraception 1996;53:115]119. w36x Bahamondes L, Marchi NM, Nakagava HM, de Melo MRL, Cristofoletti ML, Pellini E et al. Self-administration with Uniject of the once-a-month injectable contraceptive cyclofem. Contraception 1997;56:301]304.