464 E. D1CZFALUSY & B.-M . LANDGREN
A new approach to contraception, with the continuous administration of small
amounts ofprogestogens, is followed since 1972 by the World Health Organization's
(WHO) Special Programme of Research in Human Reproduction; these studies aim
at developing vaginal devices releasing the progestogens at a constant (near zero
order) rate in low quantities (WHO, 1979). The vaginal delivery systems can be left in
situ for prolonged periods of time (3 months or longer), and it is not necessary to
remove them during menstruation. These systems have not reached Phase III studies
as yet; hence their efficacy and the frequency of side effects in a sufficiently large
population remains to be determined.
The development of an improved progestogen-only contraceptive system (such as a
vaginal device) would represent a very attractive alternative to combined oral
contraceptives. Important advantages would be: few if any metabolic effects, or
minimal steroid load, and no 'first passage' effect through the liver. The problems to
be solved are how to improve the efficacy and diminish intermenstrual bleeding, and
the dilemma is how to achieve this.
Experience with other contraceptive modalities, such as long-acting injectable
formulations, indicate that increasing the dose ofa progestogen will improve efficacy,
but will also increase the frequency of intermenstrual bleeding, whereas decreasing
the amount of progestogen administered will result in a change in the opposite
How can c1inical pharmacology assist in resolving the problem? First of all by
fmding out the facts; what is the exact nature of the hormonal changes associated
with the use of progestogen-only contraceptives, what is the extent of variation in
endocrine response to a constant dose , both between subjects and within subjects,
and what is the relationship (if any) between endocrine response and bleeding
patterns? Since a review of the available literature indicated that information on
these aspects was scanty or non-existent, five years ago - with the support of the
WHO Special Programme of Research in Human Reproduction (1979) - aseries of
investigations were initiated in an attempt to provide some answers to these
Hormonal levels eharaeterizing anormalovulatory eycle
A necessary condition for the assessment of the variation in endocrine response to
low dose progestogens on the basis of peripheral hormone levels is the knowledge of
their range of levels in presumably normal, ovulatory cycles. A review of the
available information reveals that relevant data are scanty and that opinions greatly
differ as to the levels of hormonal indices in peripheral blood, which could be
considered with confidence as indicative of anormalovulatory cycle (Diczfalusy &
Landgren, 1977). It was therefore essential to establish such values in the study
population in which the progestogens were to be studied. The limits ofthe peripheral
hormone levels at various phases of the cycle which cover more than 90% of the
population studied are indicated in Table I.
An even more stringent proof of normal luteal function was provided by the
0/16 nmol Iitrerlfor a minimum offive days (Landgren, Unden & Diczfalusy, 1980).
Thirty-seven of these subjects were also followed by daily assays of the peripheral
DILEMMAS IN CONTRACEPTIVE DEVELOPMENT 465
Table 1 Peripheral hormone levels indicating normal gonadotrophin. follicular and luteal
function during various phases of the cycle. Numbers in parentheses refer to the number of
normally menstruating women studied by daily hormone assays.
l7-Hydroxy- 20a-DihydroFSH LH Oestradiol progesterone Progestetone progesterone
daysl-6 0.9-4.0 1.0-3.5 0.15-0.37 0.60-1.80 1.2-4.4 0.5-1.5
peak 2.5-16 13-36 0.70-2.10 3.2-6.3
maximum 0.48-1 .20 5.5-11.8 32-96 7.3-25
luteal phase 1.5-3.3 1.1-4.3 0.30-0.70 3.2-6.4 15-42 4.4-12
FSH and LH levels are expressed in i.u. litre! (standard: 69/104), the steroid levels are given as
nmol litre'", According to Landgren, Unden & Diczfalusy (1980) and unpublished data by
plasma levels of 17-hydroxyprogesterone and 20 a -di hydroprogesterone. Thirty-six
of the thirty-seven subjects exhibited a minimum /7-hydroxyproges/erone level of4
nmol litre:' for a minimum of five successive days and a minimum
20a -dihydroprogesterone level of 4.5 nmol litre:' for a minimum ofsix successive
permit a critical assessment ofthe endocrine response ofsubjects using different types
oflow dose progestogens for contraception.
Classifrcation of ovarian responses to low dose progestogens
As a logical next step, the endocrine effects of the 300IJg norethisterone (NET)
minipill was assessed in forty-three normally menstruating women by the daily
& Diczfalusy, 1980). This study revealed the existence of four characteristic and
distinctly different types of ovarian reaction to the NET minipill; group A exhibited
no signs of follicular or luteal activity, as evidenced by the rather constant, low
oestradiol and progesterone levels, group B showed marked cyclic follicular activity,
but no luteal function , group C had anormal follicular activity, but insufficient luteal
activity (based on the criteria shown above), whereas the oestradiol and progesterone
levels in group D were indistinguishable from those found in normally menstruating
women. Representative examples (two subjects, each exhibiting a given type of
ovarian response are presented in Figure 1.
This classification permits an adequate assessment of the endocrine response to
low dose progestogens also in most longitudinal studies, in wh ich only three blood
sampies are taken per week (Landgren, Johannisson, Masironi & Diczfalusy, 1979a).
Furthermore, the contention that the most important characteristic of anormal
luteal function is a minimum progesterone level of 16 nmol litre-' for a minimum of
five days receives further support from the results of parallel assays ofthe daily levels
of progesterone, 17-hydroxyprogesterone and 20a -dihydroprogesterone in 42
minipill-treated cycles in 21 women. Normal (D-type) luteal function was indicated
by these indices in 16, 16 and 14 cycles, respectively, and absent (types A + B) , or
insufficent (type C) luteal function was invariably found by all three indices in 26
cycles (Landgren, Lager & Diczfalusy, unpublished data).
466 E. DI CZFA LUSY & B.-M. LA NDGR EN
lOOOl pmol Iitre- I nmol litre-' [<A) «»
o :111111:n:701flllill'flillm 0
(2 nd 104 0NfH ) nmol Ir t r e" ! 40
'000 prnol r:Ire- I TREATMENT ( 0 ) (2 1'ld "-ONTH )
20001 (2 nd MON t M ) prOQesterone
(0) pmoi lltre- ' nmol li t re- I t:l
oest rudrol TREATMENT proge slerone
Figure I Characterization of the ovar ian response to low dose progestogens on the basis of
peripheral oestradiol and progesterone levels (according to Landgren & Diczfalusy, 1980). Open
circles indicate oestradiol, filled triangles progesteron e levels. Filled bars den ote bleed ing,
hatched bars spotting. Ovarian responses of different types as obtained in each of two subjects
are shown from the top to the bottom of the figure. T ype .'I: suppressed oestradiol and
progesterone levels, type B: fluctuation in oestradiol but no rise in progesteron e levels, type C:
approximately norm al oestradiol associated with insufficient rise in progesteron e levels, and
type D (bottom figure): norm al (ovulatory) oestradiol and progesterone levels.
Ovarian reaction to orally, or vaginally administered progestogens
Us ing th e classifica tion of Landgren & Di czfalusy (1 980) as descr ibed above, th e
endocrine resp on se elicited by th e NET minipill with th at ind uced by vaginal
deli very syste ms relea sing 50 IJg!24 hand 200 IJg!24 h NET respe cti vely, during a
period of 90 days, was co m pared. Segments of 30 da ys of use were a na lyzed: th e
results are shown in T abl e 2.
DIL EMM AS IN CONTRACEPTIVE DEVELOPM ENT 467
Table 2 O vari an reacti on to sma ll doses of progesto gen s ad minister ed ora lly, or by vagina l
Norethisterone l.evonorgestrel
Acc ording to Landgren & Diczfalusy (1980); O riowo , Landgren & Diczfa lusy, unpublished
(norethisterone rings); Lan dgren , Masironi & Diczfa lusy. unpublished (levo no rgestrel rings).
Preliminary results obtained with vagina l de vice s releasing levonorgestrel at a rat e
of 20 ~g/24 h are al so indicated. It is apparent from the da ta of Tabl e 2 that there
were 40 % normal (ovul atory) cycles in women taking the NET rninipill, and that in
a lmost as many subjects (39%), ovulation was inhibited (Landgren & Diczfalusy,
1980). Hence there is a marked between subject variation in the endocrine response to
It also a ppears from th e da ta of Table 2 that the ovulation inhibiting potency of
200 lJg of NET released continuously from vagina l de vices during 24 h was
considerably stro nger than that of300 ug NET given as a minipill: no luteal functi on
was ob ser ved (and thus ov ulation was certainl y inhibi ted) in 69 % of th e segments
studied in the former gro up co m pa red to 39% in th e latter. Hence it is possib le to
reduce the ste ro id load required for ov ulation inhibition by a continuous deliver y of
The data of Table 2 also ind icate th at th e ovulation inhibiting efTect of vaginal
devices releasing 50 ~g/24 h ofNET was minimal. This fmd ing is important, in view
of the fact that recently the WHO-sponsored multicentre c1inical studies with thi s
dev ice were discontinued, because several pregnancies occurred (WHO, 1979).
Whether the same will happen in the case of systems releasing 20 ug of norgestrel
rem ains to be seen; however, this approach seems to ofTer more pr orn ise, since thi s
properties of the cer vical mu cu s.
Within subj ect variation in response
In 21ofthe 43 subjects indicated in Table 2 who were taking the NET minipill, it was
possible to follow up the efTect of th e progestogen for a minimum of six months.
During the sixth treatment month, dail y blood sampi es were withdrawn, and the
hormonal profile was re-examined. Table 3 shows a comparison of the efTect of the
N ET minipill adm inistra tion during the second and sixth months in the se 21 subjects,
together with an indication of the variation in individual responses during 90 da ys
(three segments) of th e 30 women with vagina l de vice s relea sin g NET at difTerent
lt can be seen that in the eight subjects in whom D-t ype (ovulatory) steroid profiles
were seen during the second month of minip ill administrat ion , ov ulatory cycles were
found a lso during the sixth month oftreatment. Al so the 13 subjects show ing insuffi -
eiern, or absent, luteal function during the second month continued to do so during
the sixth month, although there were transitions from one type of reaction to
another (for exarnpl e, A to C, or vice versal. Although, ob viously, a grea t deal more
information is needed, these data seem to confirm the previous suggesti on , that the
468 E. DICZFALUSY & B.-M. LANDGREN
Table 3 Variation with time in the ovarian response to the 300 \Jg norethisterone minipill
(upper table) and to vaginal devices releasing norethisterone at a constant rate of 50 1Jg!24 h
and 200 g/24 h, respectively (iower table).
I. Minipill administration (continuous)
Ovarian Number of Ovarian A B C D
2. Vaginal devices: Range of ovarian response during 90 days with the device continuously
Release rate: 50~24 h* 200 ~g/24 h*
Accord ing to Landgren & Diczfalusy (1980); Landgren , Lager & Diczfalusy (to be publ ished);
Oriowo , Landgren &Diczfalusy (to be published).
significantly longer (P <0.001) and that ofthe pretreatment luteal phase significantly
shorter (P<0.001) than in the 17 subjects exhibiting normal, ovulatory (D-type)
endocrine response to the subsequent minipill treatment (Landgren & Diczfalusy,
1980). Furthermore, those subjects who exhibited no luteal activity (groups A and B)
during the second month of minipill adm inistration had significantly lower (P< 0.01)
lutein izing hormone (LH) levels in the pretreatment (control) cycles than those who
Diczfalusy, 1979b). Hence , cont inued studies ofthe predictive value ofpretreatment
hormone assays may be rewarding.
On the other hand , the data of Table 3 seem to indicate that the within subject
variability in endocrine respon se of women with vaginal devices releasing NET at a
rate of200 ~g124 h is larger than that of the subjects taking the minipilI ; of the seven
subjects with a D-type response during at least one segment , six showed other
react ions during the 90-day treatment period, which varied from type A to type C.
Furtherrnore, 12 of the 15 subjects in this group exhibited at least one segment with
DILEMMAS IN CONTRACEPTIVE DEVELOPMENT 469
""28= nmol lit re- I J'I~!~\!~~~~~"..~~
Figure 2 Peripheral plasma levels of follicle-stirnulating hormone (FSH: 0), luteinizing
hormone (RIA-LH: .), oestradiol (E2: 0), progesterone (P: e), l7-hydroxyprogesterone
(I7-HO-P: 6) and 20a-dihydroprogesterone (20a-HO-P: A) in four subjects during the sixth
month ofadministration ofthe 300 IJg norethisterone minipill. (Landgren, Lager & Diczfalusy:
470 E. DICZFALUSY & B.-M. LANDGREN
ovulation inhibition (type A and B reactions). Although these differences must be
interpreted with some caution, since the two studies (minipill versus vaginal device)
followed different designs and were not carried out simultaneously in a random
fashion, the data are sufficiently important to suggest further investigations of this
Gonadotrophin levels during low dose progestogen administration
A puzzling observation, reported recently, was the lack of any correlation between
immunoreactive follicle stimulating hormone (FSH) and both immunoreactive and
biologically active LH levels (Balogh, Robertson & Diczfalusy, 1979), on the one
hand, and the ovarian reaction during the second month of treatrnent with the NET
minipilI, on the other hand (Landgren et al., 1979b). Cycles classified on the basis of
daily progesterone levels as ovulatory exhibited just as aberrant, or flat ,
gonadotrophin profiles as anovulatory cycles, and in a few subjects there was a
Lager & Diczfalusy, unpublished). In this study, the daily progesterone assays were
complemented with the determination of 17-hydroxyprogesterone and
20a -dihydroprogesterone levels. Four representative examples are shown in Figure
The data of Figure 2 indicate that all treatment cycles were ovulatory, as evidenced
by the progesterone, 17-hydroxyprogesterone, 20 a -dihydroprogesterone and
oestradiol profiles, which fulfilled the criteria described in this paper for the
characterization of normal cycles. On the other hand, the FSH and LH profiles were
'bizarre', to say the least, and the relationship between any LH surge and subsequent
rise in progesterone, 17-hydroxyprogesterone and 20a -dihydroprogesterone levels
was tenuous, or non-existent. It appears therefore that FSH and LH assays are of
limited value in the assessment of the endocrine response to progestogens. The da ta
of Figure 2 confirm and extend our previous conclusions that ovarian suppression by
the NET minipill is unrelated to the degree of inhibition of FSH and LH secretion
(Landgren et al., 1979b) . They also raise a rather fundamental question concerning
the quantitative significance ofthe midcycle LH surge for the induction ofovulation.
Endocrine response and bleeding performance
Duringthe periods covered by hormone assays, an attempt was also made to analyze
the bleeding profiles in 38 NET minipill users and in 30 subjects using NET-releasing
vaginal devices. The da ta are compiled in Table 4.
Table 4 Relationship between the endocrine response to small dosesofnorethisterone and
percentage of days with bleeding and spotting. (Accordingto Diczfalusy& Landgren, 1981).
* Five ofthe 43 subjectsdid not keep proper menstrual diary cards.
DI LEM M AS IN CONTRACEPTIVE DEVELOPMENT 471
In th e pretreatment cyc les, 39 1 da ys ofbleedi ng out ofa total of2 121 days (18.7%)
were observed. The co rrespo ndi ng percentage of days wit h bleedin g a nd spo tting in
minipill users (254/ 1134) was 22 .3%, with th e 50 IJg/ 24 h NET vagina l devices
(274/ 1266) it was 21.5%, and with the 200 IJg/24 h NE T devices it a mo unted to 33.2%
(44 1/1329 days). lt is especia lly striki ng that in th e last gro up the numbe r ofdays with
bleeding and spo tti ng dur ing the second segment (which co rre sponds to the seco nd
month of NET administra tion) was very high (1 73/ 388) , a lmo st 45 %. T he reason for
th is is incom pletely co mprehende d. lt is obvi ous . however , th at the number of days
of bleeding an d spo tti ng of users of the 300 IJg NE T min ipill was signi ficantly less
(P < O.OOI) th an in those using vagi na l devices releasi ng 200 IJg/ 24 h of thi s
com po und. Hence, a co nsta nt co ntinuo us release of thi s particul ar progestogen
wor sen s rather than im proves th e bleeding performan ce. Aga in, th e limitat ion s ofthe
approac h em ploye d sho uld be borne in mind. Whethe r the sa me a pp lies a lso to other
c1 asses ofprogestogen s remains to be investigated.
Perh ap s the mo st im po rta nt qu estion in thi s co ntext is wheth er or not. in NET -
treated subjects, a relation ship exists between the frequ enc y of bleedi ng and spott ing,
on the one hand, and the ovari an reaction, on th e other hand. Th e data of Table 4
suggest that a relationship might ind eed exist , although it see ms to be a com plicated
Stat istical analysis ofthe data revealed that, among th e women using th e 200 IJg124
h NET devi ces, those exhibiting an ovarian reaction ofthe A - and B-t ype (lack ofan y
lut eal function and ovulat ion ) bled more frequentl y (P<0.0 5) than th ose showing
so rne lut eal activity (C- and D-t ype). However , th e type of endocri ne resp on se is
certa inl y not the onl y factor respo nsible for dilTerences in bleeding patterns, since
signi ficantly more bleeding a nd spotting (P < O.OOI) was found a mo ng wo me n with
th e 200 IJg124 h NET , who ex hibited a type A ovaria n respon se, com pa red to th ose
with th e sa me A-ty pe reacti on ind uced by th e administration of th e 300 IJg NET
minipill. Furthermore. so far no significant dilTerence in bleedi ng pattern s was found
in wornen ta king the minipilI, irrespective ofthe type of ovaria n rcac tion .
A WHO sym posium on stero idal co ntrace ptio n and mech an ism s of endo me trial
bleed ing (Diczfalu sy, Frazer & Webb , 1980) recen tl y co ncl uded th at th ere is a
pau cit y, if not lack, of info rma tion co nce rn ing th e factors resp on sible for int er- .
menst rual bleeding in wo me n using stero ida l co ntra ce ptives. Since th e relation ship
bet ween suc h bleeding and th e peripheral levels of the ad mi nistered, or endogenous,
steroi ds appea rs to be a compl ex one, it would secm esse ntial to conduct a variety of
studies at th e endometriallevel on th e local concentration and binding of endogenous
and exogenous steroids, va rio us prostaglandins and non -steroidal hormon es. Data
are urgentl y needed also on th e endometrial concentration of lysosornes, elastin, as
weil as on capillary perrneability, the de velopment of arteriol es a nd venous lacunae
in the endometrium under th e inlluence of various progestogen s, and on their
correlation with the endocr ine respon se. Although th e probl em is co m plex a nd
diffi cult, it is researchabl e, a nd syste rnatic, carefully designed and co nducted studies
sim ply ca nno t fail to provid e necessar y and mu ch needed informa tio n.
In orde r to provide a basis for the critica l assessme nt of variation in endocri ne
resp on se to low dose progestogens, the ran ge of per iph eral plasm a levels o f ovaria n
hormon es characteriz ing a no rma lovulatory cycle was established in a local study
472 E. DICZFALUSY & B.-M . LANDGREN
population . Ninety-five of hundred normally menstruating women exhibited a
minimum plasma progesterone level of 16 nmol Iitre" for a minimum of five da ys,
and 36 of 37 among them showed a Iuteal 17-hydroxyprogesterone level of at least 4
nmol litre'" for five successive days and a minimum 20a-dihydroprogesterone level
of 4.5 nmol litre"' for at least six days. These levels together with aseries of FSH, LH
and oestradiol levels indicated in the paper are considered to be characteristic for
On the basis of daily oestradiol and progesterone assays the endocrine responses of
individual subjects to small doses of progestogens were shown to fall into four distinct
categories: Type A with no signs of follicular or luteal activity, type B with marked
follicular, but no luteal activity, type C with normal follicular, but irisufficient luteal
activity, and type D with normal (ovulatory) follicular and luteal activities.
The daily hormonal profiles of43 subjects using the 300 ug norethisterone minipill
du ring a total of 1283 days were compared with those of 15 subjects using vaginal
devices with a release rate of 200 ~g124 h of norethisterone for a total of 1266 days.
The endocrine response to a given dose showed a marked variation between subjects.
In addition, the within subject variation in response appeared to be greater in vaginal
ring users than in minipill users.
The ovulation inhibiting potency of 200 ~g norethisterone released continuously
from the devices was stronger than that of the 300 ug norethisterone minipill (69%
versus 39%), indicating that the steroid load can be diminished by continuous
The length of the follicular phase in relation to that of the luteal phase in the
pretreatment (control) cycles was found to be of predictive value for the subsequent
endocrine response to the minipill. On the other hand, the degree of inhibition of
FSH and LH secretion during minipill treatment was unrelated to ovulation inhibition.
The use ofthe vaginal devices releasing norethisterone at a rate of200 ~g/24 h was
associated with significantly more bleeding irregularities than that ofthe minipilI; the
number of days with bleeding and spotting was 441 out of 1329 (33 .2%) and 254 out
of1l34 (22 .3%), respectively. In the former group, women showing no luteal function
bled more than those exhibiting some luteal activity.
ofthe intermenstrual bleeding induced by these agents.
The studies reported in this paper received fmancial support from the World Health
Organization Special Programme ofResearch on Human Reproduction.
Comments
Post a Comment
اكتب تعليق حول الموضوع