Postcoital contraception with levonorgestrel during the peri-ovulatory phase of the menstrual cycle

CONTRACEPTION

POSTCOITAL CONTRACEPTION WITH LEVONORGESTKEL DURING THE PEKI-OVULATORY PHASE OF THE NENSTKUAL CYCLE Task Force on Post-ovulatory Methods for Fertility Regulation* Special Programme of Research, Development and Research Training in Human Reproduction World Health Organization, Geneva, Switzerland

* The followiw UK

investigators participated in this study:

Ehattacharjee, Department of Obstetrics and Gynaecology, 3.K. Institute of Postgraduate Medical Education and Research, Calcutta, India

DK J. Romeo, Institute of Endocrinology and Xetabolic Diseases, Havana, Cuba Dr E.S. Kononova, Institute of Obstetrics and Gynaecology of the Academy of lledical Sciences of the USSR, Leningrad, USSR Dr A. Pretnar-Darovec, Department of Obstetrics and Gynaecology, University Medical Centre, Ljubljana, Yugoslavia Dr L. Saraya, Department of Obstetrics and Gynaecology, All India Institute of Medical Sciences, New Delhi, India i)r

Shi Yang-en, Shanghai Institute for Planned Parenthood Kesearch, Shanghai, People's Republic of China

Dr R.N.V. Prdsad, Department of Obstetrics and Gyndecology, Nation,*1 University of Singapore, !Singapore DK

G. BCrtfai, Department of Obstetrics and Gynaecology, University Medical School, Szeged, Hungary

Dr 8. Soukhria, Centre d'i3tude et de Recherfhe en Reproductioo Humaine, Tunis, Tunisia Dr P.P.A. Van Look (study coordinator), Dr D. Machin and !3s W. ZicRrian, World Health Organization, Geneva, Switzerland Kequests for reprints should be sent to Dr P.F.A. Van Look, Special Programme of Research, Development and Research Training in Human Keproduction, World Health Organization, 1211 Geneva 27, Switzerland Submitted for publication July 17, 1987 Accepted for publication September 8, 1987

SEPTEMBER

1987 VOL. 36 NO. 3

275

CONTRACEPTION

ABSTdACT The contraceptive efficacy and side effects of postcoital levonorgestrel used repeatedly during the peri-ovulatory period of one cycle was examined in 259 women. 411 subjects were of proven fertility in their present union and had ovulatory cycles as assessed from pre-treatment BBT charts. The mean number of coital acts during the treatment cycle was 7.5 (SUi2.6) and the mean number of 0.75mg levonorgestrel tablets taken during the peri-ovulatory period was 4.0 (jU:l.2). Two pregnancies, both considered to be method failures, occurred, oCJiving a fs.ilure rate of 0.8% per treated cycle. Al though the overall effect of levonorgestrel on menstrual cycle length was small and insignificant, menstrual cycle disturbances were not uncommon. Intermenstrual bleedins or spotting occurred in 8.5% of the treated cycles and 12.5% of the cycles were less than 20 or more than 35 days. Other side effects, mainly nausea, headache and dizziness, were reported by dbout 20% of the subjects but the apparent incidence of these complaints vdried markedly between tile nine participating ceritres from 0% to just over 50%. The data suggest that re,)eated postcoital use of levonorgestrel is probably not a viable approach to fertility regulation for the majority of women who have regular intercourse and wish to limit the number 3f their pregnancies. INTKOUUCTION Estrogens in tiiyh doses, estrogen-gestagen combinations and intrauterine devices (IUDs) are effective methods of postcoital contraception for tlmergency use (1). There remains an unmet need however for a regular postcoital agent that can be takeu repeatedly, if required, during the same or successive cycles. The development of such an agent has been identified as a research priority on several by the European Yedical Research occasions including, most recently, Council’s Advisory Subgroup on Human Reproduction (2), but to date no postcoitaL agent suitable for repeated use has been identified. During the 1970s the efficacy and side effects of postcoitally administered levonor&estreL have been investigated ii a number of South American trials. Uoses of levonorgestrel ranging from 0.15mg to lmg and taken within 3 hours after sexual intercourse were used in those studies. Except for the lowest dose of c).l5mg, efficacy was acceptable with method failure rates (Pearl indices) ranging from 0.7 to 3.8 (3). Nore recently, repeated postcoital use oi levoaorgestrel in a dose of 0.75mg has been assessed in a multicentre trial conducted in Hungary (4). Amongst 1,315 women (8,515 cycles) a total of 23 pregnancies occurred; six of these were considered method failares (Pearl index:O. S), the remainder user failures. Since only women with infrequent intercourse were permitted to take part in that study, the present trial was initiated to investigate the potential of this approach to postcoital contraception for women of proven fertility and at higher risk of conception.

276

SEPTEMBER

1987 VOL. 36 NO. 3

CONTRACEPTION

SUBJECTS AND METHODS Healthy, informed volunteers were recruited for this trial which had been approved by the World Health Organization’s Secretariat Committee on Research involving Human Subjects and by the corresponding local committees of the nine participating centres (see list of principal investigators for location of centres). Each of the centres was asked to recruit a total of 50 women for this Phase II trial. 1.

Criteria

for

subject

selection

Women recruited for this study were aged between 21 and 40 years, married and of proven fertility, i.e. they had been pregnant by their current husband wit’nin the last Eive years. They werr required to have a history of regular menstrual cycles (25-36 days) and to be in Subjects with a contraindication to good health and sexually active. hormonal contraception, nursing mothers and women with a history of pelvic inflammatory disease since the last pregnancy, postabortal or postpartum sepsis were not admitted to the study. Also excluded were subjects who had used hormonal contraception or an IUD during the three months prior to recruitment, who had been pregnant less than one month before and those who had abnormal findings on pelvic examination. 2.

Design -----__--

of

Tne study treatment and

the

stu*

consisted of two one post-treatment

(or three) cycle.

pre-treatment

cycles,

one

To confirm the ovulatory nature of their zycles all recruited women were asked to record their basal body temperature (BBT) during two pre-treatment cycles. Subjects who had a monophasic BUT chart during the first cycle were discontinued from the study unless the investiyator felt that the 5BT might not have been taken correctly, in which case the subject was requested to complete an additional control cycle. A third pre-treatment cycle was also required from women whose second control cycle was monophasic. Thus, only those women who had two biphasic BBT charts were eliyible to continue with the study. In both pre-treatment cycles the day of ovulation (i.e. the day prior to a temperature shift of at least 0.3%) was identified relative to the start of the cycle and, if the difference was more than four days, the subject was discontinued from the trial or an additional control cycle was obtained. If the difference was four days or less, the average was used to estimate the cycle day when ovulation was likely to take place during the treatment cycle. This estimated day of ovulation was designated day 0 and subjects were instructed to take levonorgestrel following intercourse during the peri-ovulatory period (day -4 to day +2> of that cycle. At all other tLoes throughout the study participants were asked to use a barrier form of contraception.

SEPTEMBER 1987 VOL. 36 NO. 3

277

CONTRACEPTION

Instructions for levonorsrstrel use (commercially available tablets of 0.751~ donated by Gedeon Richter Ltd., Hungary) were as After the first coitus during the peri-ovulatory period one follows. tablet was taken as soon as possible and not later than eight hours, and a second tablet 24 nours later regardless of whether Eurther coital acts had taken place during that time. Subsequently, one coitaL exposure with a maximum tablet was taken Eol.Lawing each Further of one tablet per 24 hours irrespective of the number of coital Thus, the maximum number of e:;~>osures during: that 24-hour +riod. tajlrts that could be taken was seven. Thrvul;houi approttimately end to record 3.

the study vomeii were Eoll~>wed-up at intervals of the dHT and vaginal bleeding four weeks to examine any conplai-its the subjects had.

Statistical ~------

chart

methods

cycle Wds Jeter.uined using ‘i?.~e pregnanc;r rate during the treatment Pearl i.ndex (number of pregnancies per 100 woman-years of In view of the v.ariation i,? length of the treatment cycle ex;)oSure). (ses Results) the actual length (in days) of the treatment cycle was us& for each subject when calculating total exposure for tile wtlole In the 3 women fur whom the actlual lrn~th of the group. study len?:th of their pre-trtiatment t raatlnc,lt i-yc le was unitflown, tllrt averdse The cycles was used .is dn estimate of treatment cycle length. confidence i;lterv.dl for ihe Pearl -i,ziex was calculated Ilsins tables of confidence intervals for the Poisson distribution (Documenta Geigy, 7th Cdi:ion, p137). the

Coraparisons ‘Jetwcen pre -treatment pre-Lreatment and post-treatment cycle paired t-test. Valaes deviations.

;ive;l

i*l br;ickets

afLer

dnd treatment, lengths were

,nean vallles

and hetdeen made by the

represent

standnrl

BESULTS .Iz total of 372 women were recruited i:lto the trial (three of the tier< unahIe to reach the recruitnent tar&et of participating centres Two-hundred-and-fifty-nine (69.6%) of these women 50 subjects). The renainiqg satisfied all the criteria specified by the protocol. the reasons shown 113 (30.4%) ulere excluded Erom the study group for in Table I. As i,ldicated in Taulc I 3J oE the 372 rtcruized wozen (8.11) werr discontinued because their cycles were anovulatory as judged from BBT a similar percentage of women was lost to Eollow-up. CIldfYt-3, and Other reasons f!~r discontinuation were Less common and amounted to 13leven of the subjects who did not l&.2% of recruited subjects. fulfil the grotocol criteria and were erroneously recruited and The reasons for were excluded during data analysis. treated,

278

SEPTEMBER

1987 VOL. 36 NO. 3

CONTRACEPTION

Table

Reason

for

I:

Subjects

excluded

exclusion

to treatment cycle uonophasic BBT loss to follow-up personal reason pregnant Jifference of more than 4 days between estimated days of ovulation medical reason other and unknown After treatment cycle protocol violations

from

study

Number

group

of

women

Prior

exclusion or more cytology subjects.

30 30 16 11

7 3 5 11

in this group were: last pregnancy Less than one month (n=Z) than five years (n=8) before recruitment and abnormal cervical There were no pregnancies in this goup of 11 (n=l> .

baseline characteristics of the 259 women who completed are given in Table II and were not different from those of women who were discontinued and/or excluded from analysis.

the study the 113

Two pregnancies occurred during the 2.59 treatment cycles for a failore rate of 3.3% per treated cycle. In both cases the pregnancy was classified as a method failure. When expressed as Pearl index the failure rate was 10.0 with 95X confidence interval 1.2 to 35.0. Complete intercourse data during pre-treatment cycles and treatment cycle were available for 213 of the 259 women. The mean number of reported coital acts during the treatment cycle was 7.5 (2.6) and slightly higher than the average of 7.0 (2.1) during the pre-treatment cycles. This difference persisted after adjusting for variation in cycle length. The number of levonorgestrel tablets taken during the per-i-ovulatory period of the treatment cycle ranged from zero (one woman) to seven (10 women) with a mean of 4.0 (1.2) tablets. The two women who became pregnant used 3 and 4 tablets, respectively. The mean length of the two pre-treatment cycles in the 259 subjects was 28.6 (2.5) days. The corresponding values for the treatment and post-treatment cycle were 25.4 (6.4) and 28.2 (4.0) days. The mean paired difference between the average Length of the pre-treatment cycles and the length of the treatment cycle (Figure 1)

SEPTEMBER

1987 VOL. 36 NO. 3

CONTRACEPTION

Table

II:

Baseline

characteristics

Study

group

Exclusions

---Number

of

women

113

259 29.6

(4.5)

28.3

(5.3)

(cm) SD)

159.5

(6.2)

159.8

(5.9)

(kg) SD)

56.3

(3.1)

56.9

(8.0)

13.7

(21.0)

14.9

(29.5)

Age (yrs) (mean;

SD)

Height (mean; Weight (mean;

Months since Last pregnancy (geometric mean; range)

interquartile

Outcome of last pregnancy (number of women, %) - live birth - spontaneous abortion - induced abortion - stillbirth

137

(53)

58 (51)

2 (1) 120 (46)

3 (3) 50 (44) 2 (2)

The was 0.2 days (Y5% confidence interval -1.0 and +0.5 days). corresponding value for the paired differences between the average length of the pre-treatment cycles and the length of the with 95% confidence interval -1.0 post-treatment cycle was 0.4 days, Thus, the overall effect of levonorgestrel on mean to +0.2 days. menstrual cycle length was small and statistically not significant. of the treatment as evidenced by Figure 1, the length However, the post-treatment cycle was affected and to a lesser extent, cycle, For example, 12.5% of the treatment in a proportion of the subjects. cycles were less than 20 or more than 35 days as compared to only 1.3% In addition, 8.5% of the women reported of the pre-treatment cycles. the appearance of intermenstrual bleeding or spotting during the As shown in Table III, there was a high incidence of treatment cycle. intermenstrual bleeding or spotting episodes in women with treatment cycles of more than 43 days.

280

SEPTEMBER

1987 VOL. 36 NO. 3

CONTRACEPTION ~~~ ~~-

14 TREATMENT

a

13 1

AVERAGE

-

1

CYCLE LENGTH MINUS

PRE-TREATMENT

CYCLE LENGTH

11 12 I 10 t 9L 1

8

? fJ 7 I 6 8 6-5’ 4321;

a,

OL

-10

-15

+20

14r~----13-

I IZ-

b

E3 Jm t25

+3f)

1~ t

POST-TREATMENT CYCLE LENGTH MINUS AVERAGE

PRE-TREATMENT CYCLE LENGTH

I1 IO 9_

Figure length average cycle.

1: Distribution of paired of the pre-treatment cycles length of the pre-treatment

SEPTEMBER

differences and the cycles

1987 VOL. 36 NO. 3

between a) the average treatment cycle, and b) the and the post-treatment

281

CONTRACEPTION

Table

III: in

Incidence relation

of intersenstrual bleeding or to length of treatment cycle

-------less

---than

_ _-.-----

----~

?T.+3 5

~__

-~1

(8.3)

224

14

(6.3)

11

L

(9.1)

36-42

incLuding was not

-12

30

* In 3 cycles, tr~atlnrnt cycle

Number (Z, of cycles with intermenstrual Iblc2dint: or spottinS

timber of cycles*

length (days)

Cycle

spotting

the tuo Xno:Jri.

conception

cycles,

the

length

of

the

?Jeither the lCength of tile treatment cycle nor the presence of intemenstrual bleeding were related to the number of tablets taken. There was also no rel;at ionship betwee;> the nnnbe: of tablets iaigested Out of the 259 women Lreated, 205 and the reporting of side effects. (79. 2%) reported no side ef fee ts wher%*as the remainin; 54 (20.8%) ilad one Jr Inure comi3laints (Table IV). Table -

IV:

Uumber --

(X)

of

women ___- reporting

side

effects-

--

____--___-____-_

-~__-___-Nausc.1

Headache Dizzi le:;s Lower -1bdoninal pain/ backache Sreast tenderness/ f n&0 rs eme n t Vo.ni ting I)t he CS

29 (11.2) 14 (5.4) 14 (5.4) 7

(2.7)

4 2

(1.5) (0.8) (3.1)

a

The number of women reporting side effects varied mclrkedly between In two of the nine centres none of the subjects reported any centres. ,side effect whereas i.n two other centres just over half of the floraen These last two centres complained of one or ,nore side effects. accounted for 57.4% of the 54 women reporting side effects and fur 66.61 of the 78 repor:ed complaints.

282

SEPTEMBER

1987 VOL. 36 NO. 3

CONTRACEPTION

In order to attain a greater degree of safety in hormonal fertility regulation it would seem desirable to limit as much as possible the amount of contraceptive drugs to which women are exposed The availabil.ity of a safe, during their reproductive years. acceptable and effective method of postcoital contraception that could would go some way towards achieving be used ad libituu if required, this objective. The development of such a method is proving to be extremely difficult, hovrever, no doubt because the properties which a drug must possess in order to be an effective postcoital adent are qu; te formidable ( 1). Levonor~e.strtiL which probabLy acts by readeril% the endometriun unsuitable ;or im?;antation (1)) has been employed as a postcoital contraceptive in emer;ency situations (5) as :lnce Erom the fourth day beEore until the second day after the estimated day of ovulation which is the period of highest fertility. In couples not using contraception, intercourse durin:: that time cdrries an average risk of conception of 18% (8). In common with other Eorms of yb>estagen-only contraception the postcoital use of levonurgestrel was associated uTith a relatively incidence of menstrual disturbances. Lntermenstrual olecding or spotting were observed in 8.5% of the treated cycles and 12.5% of

SEPTEMBER 1987 VOL. 36 NO. 3

high the

283

CONTRACEPTION

cycles were less than 20 or more than 35 days. The high incidence of intermenstrual bleeding or spotting episodes in women with treatment cycles of more than 43 days (Table III) suggests that these episodes were most probably instances of aberrant menstruation rather than true intermenstrual bleeding. The overall incidence of cycle abnormalities in the present study was somewhat lower than the 27.5% reported by SeregQly (4) and markedly lower than the incfdence of 88.9% observed by Canzler et al. (9) who used the same dose of 0.75mg levonorgestrel -in 27 women during 226 cycles with 1 method failure. The high incidence of cycle irregularities (mostly breakthrough bleeding and oligonenorrhoea) in that study may have been due to the young age of the study subjects (mean: 16.8 yrs, range: 15-19 yrs). The neuroendocrine mechanisms underlying regular ovarian function are generally considered to be less wall established in this age group as evidenced by the greater variability in menstrual cycle length (10) and the higher susceptibility of such women with irregular cycles to post-pill amenorrhoea (11). The incidence of reported side effects varied markedly between the In two of the centres (New Delhi and Tunis) participating centres. over half of the treated subjects had one or more conplaints just whereas no or few side effects were reported by the women in other The reason for this large variability is not known. It may centres. be due to differences in the questioning technique used by the investigators involved. Alternatively, the observed variation may be In this context it is worthy of note that the two culturally related. centres with the highest incidence were situated in countries where a high level of user acceptance, as family planning has not yet gained evidenced by the higher fertility rates and lower rdtes of contraceptive prevalence compared to the other countries represented in this study (12). dhen comparing Pearl indices the effectiveness of postcoital levonorgestrel observed in this study is comparable to that of bdKrieK forms of contraception but substantially lover than that of other It should be noted however that a f onus of hormonal contraception. comparison based on the Pearl index is not entirely appropriate because two different contraceptive Imethods (postcoital levonorgestrel and barrier methods) were used during the treatment cycle in the present study. Also, the selection criteria employed, in particular the requirements of proven fertility and MT evidence of ovulation, are likely to have created a more fertile population than those usually recruited for effectiveness studies of other contraceptive Nevertheless, methods. repeated postcoital use of levonorgestrel does not appear to be a viable approach to Fertility regulntion for the majority of women of reproductive age who have regular intercourse and wish to limit the number of their pregnancies. For such women the total drug load per cycle would be greater than during use of a gestagen-;nly pilL and the incidence of cycle disturbances might be higher than with other forms of gestagen-only contraception because of intermittent use of the drug. Postcoital. coritraception the irregular,

SEPTEMBER

1987 VOL. 36 NO. 3

CONTRACEPTION may be of value however to certain women such as those needing contraception for a short period only and/or havine intercourse infrequently. The failure rates per treatment cycle observed in this and other studies (4,5,9) suggest that postcoital administration of 0.75mg levonorgestrel might also represent an effective alternative to high dose estrogens or estrogenlgestagen combinations for use in emergency situations. Further studies are required however to validate these suggested uses. ACKNOWLEUGEMENTS Levoaorgestrel was kindly donated by Gedeon Richter Ltd., The assistance of MS N. Laperriere who participated in data Hungary. processing and analysis and of Xrs J. &rozzi who typed the manuscript is gratefully acknowledged.

‘\dFEKENCES contraception: a cover-up story. and Tomorrow (Diczfalusy, E. and Wew York, pp 29-42 (1987) Press,

In:

1..

Postcoital V
2.

European Yedical Research Council’s Advisory Subgroup on Human Recommendations for priority areas in human Reproduction. reproduction research, 1984. Lancet i: 1228-1230 (1984)

3.

Rinehart, Reports,

4.

Sereggly, Hung. 30:

5.

Postcoital contraception: experiences with Hoffmann, K.O.Y. ethinyloestradiol/norgestrel and levonorgestrel only. In: Fertility and Sterility (Harrison, R.F., Bonnar, J. and Thompson, W. , Eds). ?ITP Press Ltd, Lancaster, U.K., pp 311-316 (1983)

b.

INring, Reprod.

7.

The menstrual VolIman, K.F. Obstetrics and Gynaecology, Philadelphia, U.S.A. (1977)

8.

A prospective multicentre trial of the World Health Organization. [II. Characteristics ovulation method of natural family planning. Fertil. Steril. of the menstrual cycle and of the fertile phase. 4J: 773-778

9.

Canzler, E., Ahrendt, H.-J. and Ahrendt, S. Levonorgestrel zur postkoitalen Kontrazeption. 1182-1191 (1984)

contraception W. Postcoital Series J: 141-156 (1976) Results G. 72-75 (1982)

G.K. Fertil.

of

a multicentre

- an appraisal.

trial

The incidence of anovular Suppl. 6: 77-81 (1969) In: cyc le. Vol. 7. W.B.

SEPTEMBER 1987VOL. 36NO. 3

cycles

of

Population

Postinor.

in women.

Ther.

J.

Major Problems in Saunders Company,

Erfahrungen mit Zbl. Gynlkol.

106:

285

CONTRACEPTION

10. Treloar, A.E., Boynton, R.E., aehn, B.G. and Brown, B.M. Variation of the human senstrual cycle through reproductive life. Int. .J.Fertil. 12: 77-126 (1967) 11. Weisberg, E. Fertility after discontinuation of oral contraceptives. Clin. Reprod. Fertil. 1: 261-272 (1982) 12. International Bank for Recoastruction aLId Development/The World dank. Population Change and Economic Development, Oxford University Press, Oxford, U.K. (1984)

286

SEPTEMBER

1987 VOL. 36 NO. 3