A comparison of the inhibition of ovulation achieved by desogestrel 75 µg and levonorgestrel 30 µg daily

C.F. Rice1,3, S.R. Killick1, T. Dieben2 and H.Coelingh Bennink2

1 Department of Obstetrics and Gynaecology, The Princess Royal Hospital, Saltshouse Road, Hull HU8 9HE, UK and 2 Clinical Development Department, NV Organon, Oss, The Netherlands


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
A randomized, double blind, group comparative study was performed over a 12 month period to compare inhibition of ovulation during the use of two (progestogen-only) oral contraceptives containing doses of 75 µg desogestrel or 30 µg levonorgestrel. Seventy-one female volunteers with regular cycles and established ovulation by ultrasonography and serum progesterone concentrations were recruited from an out-patient clinic in a university hospital and asked to participate in the study. Transvaginal ultrasonography and serum oestradiol, progesterone, luteinizing hormone (LH) and follicle stimulating hormone (FSH) measurements were performed throughout the 7th and 12th 28 day treatment period. Desogestrel at a dose of 75 µg showed a significant inhibition of ovulation compared to 30 µg levonorgestrel (P < 0.001).

Key words: desogestrel/inhibition of ovulation/levonorgestrel/progestogen-only contraceptives


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Progestogen only pills (POP) have been used in the UK for many years, but still have a limited acceptability by both women requesting contraception and doctors prescribing them. They account for only 8% of the current UK oral contraceptive market. The current POP contain levonorgestrel, norethisterone or ethynodiol diacetate. Their mode of action is multi-faceted and includes ovulation inhibition, mucus hostility, endometrial changes and alteration in the motility of the Fallopian tube. The failure rate of the POP (0.7–1.8/100 woman–years) is higher than that of the combined contraceptive pill (0.17–0.41/100 woman–years) (Oxford/FPA Study, 1982Go). Although the need for timing of tablet ingestion, and hence patient compliance, contributes to the failure rate of both preparations, ovulation is not always inhibited with the POP, which may lead to the higher method failure.

Previous studies (Skouby, 1976Go; Viinikka et al., 1976Go; Rice et al., 1996Go) have shown that when desogestrel is administered at a dose of 60–75 µg/day, ovulation is inhibited completely. This suggests that it would be a more effective POP than levonorgestrel 30 µg/day, which prevents ovulation in only 40% of cycles and relies upon the other factors mentioned above for its contraceptive abilities. In this paper, the ability of desogestrel to inhibit ovulation in healthy female volunteers was compared with a currently available preparation of levonorgestrel.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Subjects
Seventy-one healthy women were recruited and randomly allocated to receive either desogestrel 75 µg/day or levonorgestrel 30 µg/day, for 13 treatment periods of 28 days (364 days). Randomization was achieved by assigning a code number (0001–0071) to the subjects, in order of inclusion: i.e. the first subject entering the study was assigned to medication bearing the lowest number, the second subject the next number, and so on. Each code number on the randomization list corresponded to either desogestrel 75 µg or levonorgestrel 30 µg treatment. The same code numbers were used to further identify subjects and their treatments, blood samples and documents. All volunteers gave informed written consent. The subjects were between 18 and 40 years, in good physical health, with normal ovulatory cycles with a mean length between 24–35 days. Body weight had to be between 80 and 120% of ideal body weight (Metropolitan Life Insurance tables). None of the subjects had used injectable hormonal methods of contraception for the past 6 months or other hormonal contraceptives for the two menstrual cycles prior to this study. Ovulation was confirmed before the start of the study by ultrasound, serum progesterone concentrations in the ovulatory range, or both.

Study design
A full medical history was taken and the volunteers then underwent a clinical examination including a pelvic examination. On day 10 of their menstrual cycle the women had a transvaginal ultrasound scan using an Ultra-mark 4-Plus (ATL) to measure follicular diameters. Both ovaries were assessed and every follicle above 5 mm, as a mean of two diameters (transverse and longitudinal), was recorded. The scans were performed on alternate days until the dominant follicle reached 15 mm in diameter. At this stage the scans were performed daily until follicular rupture (Queenan et al., 1980Go). Blood was taken at each visit and analysed for 17ß oestradiol, luteinizing hormone (LH) and follicle stimulating hormone (FSH) by Microparticle Enzyme Immunoassay (Imx assay; Abbott Laboratories, Diagnostic Division, Abbott Park, IL, USA). At 4, 7 and 10 days post-ovulation, blood was taken for analysis of progesterone concentrations by solid phase fluorimmunoassay (Delfia assay; Wallac UK Ltd, Milton Keynes, UK) in addition to oestradiol concentrations. The women who had demonstrated ovulatory cycles were asked to commence medication on the first day of their next menstrual cycle. The tablets were taken daily with no break in between packets.

After three treatment periods, the women returned for a check to measure weight and blood pressure and to record any changes in physical well being, in addition to the compliance to the medication.

The women were evaluated throughout the 7th and 12th treatment period. During each of these 28 day treatment periods they attended twice a week and underwent an ultrasound examination to assess any follicular development. When a follicle greater than 15 mm was measured, the scans were performed daily until follicular rupture or until the measurements had been static for 3 consecutive days. The scans were then continued twice a week. A blood sample was taken for analysis of serum progesterone, oestradiol, FSH and LH concentrations at every visit.

Statistics
When testing at the usual 5% level, a sample size between 30 and 40 subjects per group would allow for an 80% chance of detecting a difference of 6–7 mm in mean maximum follicular diameter and a difference of 10% in the desogestrel group versus 40–47% in the levonorgestrel group with respect to the incidence of progesterone concentrations >=10 nmol/l.

The statistical analysis was carried out on the intent-to-treat groups. The treatment groups were compared for maximum progesterone values using the {chi}2 test. Maximum FSH and LH values and maximum and mean oestradiol values were analysed using a Wilcoxon rank sum test.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
After the initial visit 71 women were randomized, seven were found to be unsuitable during screening due to anovulation, cyst formation or a prolonged menstrual cycle. The 64 eligible women started the medication on the first day of their next menses. The disposition of the volunteers is shown in Table IGo.


View this table:
[in this window]
[in a new window]
 
Table I. Disposition of volunteers
 
Seven women failed to complete the full year of medication. The majority withdrew because of adverse side effects: one in the desogestrel group for depression, fatigue and headache, one in the levonorgestrel group for pre-menstrual tension, and the other three for bleeding disturbances (two in the desogestrel group and one in the levonorgestrel group). One woman developed an intercurrent illness (desogestrel group), and another moved abroad during the trial (levonorgestrel group). The groups were comparable at their initial assessment (see Table IIGo).


View this table:
[in this window]
[in a new window]
 
Table II. Patient characteristics
 
Follicular activity and maximum progesterone values
During the screening cycle 64 women demonstrated ovulation on ultrasound scan. This was defined as follicular growth above 15 mm diameter with subsequent rupture (a decrease in the diameter of the follicle by >50% within 24 h) (Hackeloer et al., 1979Go).

Ultrasound observations from the two assessment periods were used to rank the degree of ovarian suppression achieved by each preparation at both 7 and 12 months. Decreasing rank order was taken to be from no follicular activity to a persistent follicle <=30 mm (Queenan et al., 1980Go) to follicular rupture. A persistent follicle was defined as a follicle between 15 and 30 mm in size which remained static in size for 3 successive days. Results are shown in Table IIIGo. In treatment period 7, no follicular activity was seen in five subjects in both treatment groups, though in treatment period 12 this had increased to nine subjects in the desogestrel group and decreased to three in the levonorgestrel group. Follicular rupture was seen in 19 cycles in the levonorgestrel group compared to only three in the desogestrel group. The number of persistent follicles was greater in the desogestrel group.


View this table:
[in this window]
[in a new window]
 
Table III. Follicular activity per treatment group and treatment period
 
In addition to the above ovarian activity, one woman in the desogestrel group had a co-incidental cyst in the 12th treatment period. In the levonorgestrel group, extra cysts were noted in three and four women in the 7th and 12th treatment period respectively.

At screening, all the subjects showed follicular rupture on ultrasound and all but three subjects had progesterone levels >30 nmol/l, confirming ovulation. In both the 7th and 12th treatment period a comparison of ovarian suppression was also made using serum progesterone levels. Decreasing rank order was taken to be from progesterone values <10 nmol/l, to a value 10–30 nmol/l to levels >30 nmol/l. The difference between the desogestrel and levonorgestrel groups for a maximum progesterone level >30 nmol/l is statistically significant (P < 0.001) for both assessment periods (see Table IVGo).


View this table:
[in this window]
[in a new window]
 
Table IV. Maximum progesterone concentrations per treatment group and treatment period
 
When the ultrasound findings were combined with the progesterone results, a more accurate indication of ovulation was obtained. This is defined as follicular rupture followed by a rise in the serum progesterone level to >30 nmol/l. In the cycles studied, levonorgestrel was found to have a greater percentage of ovulatory cycles. The results for each treatment group (both assessment periods combined) are given in Table VGo. The results from both treatment periods were combined for each treatment group.


View this table:
[in this window]
[in a new window]
 
Table V. Follicular rupture and maximum serum progesterone concentrations per treatment group in the cycles studied
 
The remaining six cycles studied (all in the levonorgestrel group) where the maximum progesterone value was >30 nmol/l showed the formation of a persistent follicle in five cycles and no follicular activity in the sixth. In the subjects who demonstrated persistent follicles or cysts, none of the desogestrel group had raised progesterone values (>10 nmol/l); whereas in the levonorgestrel group, seven out of 18 women (39%) with persistent follicles, and five out of 12 women (42%) with cysts, had raised progesterone values, i.e. luteinized unruptured follicles (LUF). None of the persistent follicles nor cysts caused any clinical symptoms. They resolved spontaneously, despite continuously high progestogen concentrations.

Maximum LH and FSH concentrations
The maximum LH levels were significantly lower in the desogestrel group during treatment cycles 7 and 12, compared to the levonorgestrel group (P < 0.001 and P = 0.003 respectively). The difference in the FSH values between the two groups was not significant in either treatment cycle (P = 0.2 and P = 0.9). See Table VIGo.


View this table:
[in this window]
[in a new window]
 
Table VI. Maximum luteinizing hormone (LH) and follicle stimulating hormone (FSH) concentrations per treatment group
 
Maximum and mean oestradiol values
The maximum oestradiol values during both treatment periods 7 and 12 were significantly lower in the desogestrel group than in the levonorgestrel group (P = 0.049 and P = 0.001, respectively). In addition to suppression of the maximum values, the mean oestradiol levels in both treatment cycles were significantly lower in the desogestrel group (P = 0.044 in cycle 7 and P = 0.003 in treatment cycle 12). See Table VIIGo.


View this table:
[in this window]
[in a new window]
 
Table VII. Maximum and mean oestradiol concentrations per treatment group
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
This study was designed to compare a new 75 µg desogestrel-only preparation against a 30 µg levonorgestrel POP which is commercially available. We have concentrated solely on ovarian activity and the inhibition of ovulation, since this is the most important mode of contraceptive action. In the present study we followed follicular development over two periods of 4 weeks at treatment periods 7 and 12. This, accompanied by serum oestradiol, progesterone, LH and FSH measurements provided a comprehensive insight into ovarian function (including ovulation) during the use of the two POP.

Follicular rupture alone does not always signify fertility and similarly a raised serum progesterone level taken in isolation is not a definitive test, as it will include those cycles with LUF. Tayob et al. (1985) used ultrasound scanning alone to examine the effects of POP on follicular activity and the majority of earlier trials on progestogens, including desogestrel, have only studied endocrine parameters with regard to ovulation (Skouby, 1976Go; Viinikka et al., 1976Go). Therefore in this study ovulation was defined as follicular rupture on scan followed by a rise in serum progesterone value. Using both criteria the desogestrel preparation demonstrated a significantly greater effect on the inhibition of ovulation. Only one subject ovulated in the desogestrel group during the 59 cycles studied. In the levonorgestrel group by comparison, 16 out of 57 cycles (28%) were shown to be ovulatory. These results strongly suggest that the desogestrel POP should be more reliable as a contraceptive agent.

The decrease observed for LH in both study periods was significantly greater in the desogestrel group indicating that with desogestrel the suppression of the hypothalamic–pituitary axis is more pronounced. No difference was found for FSH. A similar number of women in both desogestrel and levonorgestrel groups formed persistent follicles.

However, the endocrine picture was very different between the two preparations. None of the women with persistent follicles or cysts in the desogestrel group showed any evidence of luteal activity compared to a 40% (12/30) incidence of LUF in the levonorgestrel group. None of the subjects required medical intervention for follicular cysts.

The maximum and mean oestradiol levels were significantly lower in the desogestrel group compared to the levonorgestrel group. Hypo-oestrogenism was not observed in this study population and all the oestradiol levels were above those which would lead to concern with regard to osteoporosis (Mehta, 1993Go).

Our findings on the levonorgestrel POP are in full agreement with earlier studies, i.e. that the primary mechanism of action of the levonorgestrel POP used in this study is not ovulation inhibition (McCann and Potter, 1994Go), but an abnormal luteal phase and inadequate cervical mucus contribute mainly to the effectiveness.

Desogestrel produced a significant inhibition of ovulation in comparison to levonorgestrel. This was illustrated by both the lack of follicular development and the suppression of endocrine values. Desogestrel, at a dose of 75 µg daily, is therefore expected to have a lower failure rate than the currently available POP.


    Acknowledgments
 
Martin Struijs is kindly acknowledged for data analysis and presentation of the results. This project was supported by a research grant from NV Organon, Oss, The Netherlands.


    Notes
 
3 To whom correspondence should be addressed Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Hackeloer, B.J., Fleming, R., Robinson, H.P. et al. (1979) Correlation of ultrasonic and endocrinologic assessment of human follicular development. Am. J. Obstet. Gynaecol., 135, 122–128.[ISI][Medline]

McCann, M.F. and Potter, L.S. (1994) Progestin-only contraception: a comprehensive review. Contraception, 50 (Suppl. 1), S13–S21.

Mehta, S. (1993) Bone loss, contraception and lactation. Acta Obstet. Gynaecol. Scand., 72, 148–156.[ISI][Medline]

Oxford/FPA (1982) Study report. Vessey,M., Lawless,M., Yeates,D. Efficacy of different contraceptive methods. Lancet, 10, 841–842.

Queenan, J.T., O'Brien, G.D., Bains, L.M. et al. (1980) Ultrasound scanning of the ovaries to detect ovulation in women. Fertil. Steril., 34, 99–105.[ISI][Medline]

Rice, C.F., Killick, S., Hickling, D. and Coelingh Bennink, H. (1996) Ovarian activity and vaginal bleeding patterns with a desogestrel-only preparation at three different doses. Hum. Reprod., 11, 101–104.

Skouby, S. (1976) The influence on the pituitary–ovarian function, cervical mucus and vaginal cytology of a new progestational compound. Contraception, 14, 529–539.[ISI][Medline]

Tayob, Y., Adams, J., Jacobs, H.S. and Guillebaud, J. (1985) Ultrasound demonstration of increased frequency of functional ovarian cysts in women using progestogen-only contraception. Br. J. Obstet. Gynaecol., 92, 1003–1009.[ISI][Medline]

Viinikka, L., Ylikorkala, O., Nummi, S. et al. (1976) Biological effects of a new and potent progestogen. A clinical study. Acta Endocrinol., 83, 429–438.[ISI][Medline]

Submitted on June 1, 1998; accepted on November 30, 1998.