Phillips et al.
to be a selective progestin on the basis of reduced androgenicity. REFERENCES l. Wahl P, Walden C, Knopp R, et al. Effect of estrogen/progestin potency on lipid/lipoprotein cholesterol. N Engl] Med 1983;308:862-7. 2. Gaspard U]. Metabolic effects of oral contraceptives. AM ] OBSTET GYNECOL 1987;157:1029-4l. 3. Fotherby K. Oral contraceptives, lipids and cardiovascular disease. Contraception 1985;31 :367 -94. 4. Wynn V, Niththyananthan R. The effect of progestins in combined oral contraceptives on serum lipids with special reference to high-density lipoproteins. AM] OBSTET GyNECOL 1982:142:766-72. 5. Chapdelaine A, Desmarais ]-L, Derman RJ. Clinical evidence of the minimal androgenic activity of norgestimate. Int] Fertil 1989;34:347-52. 6. McGuire ]L, Phillips A, Hahn DW, et al. Pharmacologic and pharmacokinetic characteristics of norgestimate and its metabolites. AM] OBSTET GYNECOL 1990; 163:2127-3l.
October 1992 Am J Obstet Gynecol
7. Phillips A, Demarest K, Hahn DW, et al. Progestational and androgenic receptor binding affinities and in-vivo activities of norgestimate and other progestins. Contraception 1990;41 :399-410. 8. Phillips A, Hahn DW, Klimek S, McGuire]L. A comparison of the potencies and activities of progestogens used in oral contraceptives. Contraception 1987;36:181-92. 9. Phillips A. The selectivity of a new progestin. Acta Obstet Gynecol Scand 1990;69:(suppl 152):21-4. 10. Killinger ], Hahn DW, Phillips A, et al. The affinity of norgestimate for uterine progestogen receptors and its direct action on the uterus. Contraception 1985;32: 31l. Il. Phillips A, Hahn DW, McGuire]L. Relative binding affinity of norgestimate and other progestins for human sex hormone-binding globulin. Steroids 1990;55:373-5. 12. EI Makhzangy MN, Wynn V, Lawrence DM. Sex hormone binding globulin capacity as an index of oestrogenicity or androgenicity in women on oral contraceptive steroids. Clin Endocrinol 1979; 10:39-45. 13. Victor A, Weiner E, Johansson EDB. Relation between sex hormone binding globulin and d-norgestrel levels in plasma. Acta Endocrinol 1977;86:430-6.
A norgestimate-containing oral contraceptive: Review of clinical studies Michael Edwin Kafrissen, MD, MSPH Raritan, New Jersey Despite the well-documented efficacy and safety of low-dose oral contraceptives, the development of newer formulations containing highly selective progestins with minimal or no androgenic activity has been a goal of pharmaceutical research. The efficacy and safety of norgestimate, a progestin with inherently low androgenicity, in combination with ethinyl estradiol, has been examined in several phase II and phase III clinical studies, and these are reviewed. Norgestimate/ethinyl estradiol has proved to be a low-dose oral contraceptive with high selectivity that provides the cycle control of older oral contraceptive formulations with comparable efficacy. Results of comparison studies between norgestimate/ethinyl estradiol and formulations containing norgestrel, a progestin with relatively greater androgenic activity, in combination with ethinyl estradiol, are reported for effects on lipid and lipoprotein levels and carbohydrate metabolism. Norgestimate/ethinyl estradiol consistently produced statistically significant increases in high-density lipoprotein cholesterol and improvement in the ratio of low-density lipoprotein cholesterol to high-density lipoprotein. In contrast, norgestrel/ethinyl estradiol produced statistically significant decreases in high-density lipoprotein cholesterol and potentially adverse changes in the low-density/high-density lipoprotein ratio. Phase II studies have confirmed that norgestimate/ethinyl estradiol has low androgenic activity and causes minimal effect on coagulation factors and carbohydrate metabolism. (AM J OSSTET GVNECOL
1992;167:1196-202.)
Key words: Cilest, Ortho-Cyclen, low-dose oral contraceptives, norgestimate, low androgenicity
From The R. W. Johnson Pharmaceutical Research Institute. Reprint requests: Michael Edwin Kafrissen, MD, MSPH, Senior Director, Clinical Reproductive Research, R. W. Johnson Pharmaceutical Research Institute, Route 202 South, Raritan, NJ 088690602. 6/0/33746
1196
With reversible, safe, and effective birth control as the objective, oral contraceptives (OCs) have been studied more widely than any other pharmacologic agents. An important step in their development was the intro-
Volume 167 "'umber 4, Part 2
duction of low-dose, fixed estrogen/progestin combinations, which dramatically reduced the incidence of adverse reactions associated with early preparations. Lowering the estrogen content to one fifth of the original dose sharply decreased the risk of thromboembolic events; lowering the progestin content resulted in substantive improvement in metabolic profiles. Currently, in healthy women under 30 years of age and in nonsmokers up to menopause, OC use provides effective birth control associated with few problems compared with those of an unintended pregnancy and affords certain health benefits as well. Recent efforts have focused on fine-tuning OCs to minimize their contribution to risk factors associated with the development of cardiovascular disease. In early studies of OCs, it became evident that the various progestins used may have androgenic and estrogenic activity as well as progestational activity. It is the effects of progestins on lipoprotein metabolism that are most readily linked to possible increased long-term cardiovascular risk with OC use: progestins tend to increase low-density lipoprotein (LDL) cholesterol and decrease high-density lipoprotein (HDL) cholesterol, including the cardioprotective HDL2 • More pronounced effects are seen with progestins that have greater androgenic activity. 1.2 Their adverse effects are most pronounced on HDL, HDL2 , and apoprotein A-I. Carbohydrate metabolism is also sensitive to the androgenicity of OC progestins. Fixed-dose, combined OCs may increase insulin resistance, resulting in elevated plasma glucose and plasma insulin levels in response to a glucose load depending on the potency of the progestin and the dose given.' Insulin resistance is another potential contributor to long-term cardiovascular risk. In general, low-dose OCs appear to affect coagulation minimally; anticoagulant and procoagulant effects have been noted. Other factors have substantial effects on coagulation, such as smoking, obesity, family history, and blood type.' However, current use of OCs remains a risk factor for acute-event cardiovascular risk. The prudent choice of OC remains a low-dose, low androgenicity formulation that is compatible with.good safety, efficacy, and tolerability. Development of a selective progestin with a targeted response and minimal androgenicity would be a step forward in efforts to eliminate the adverse metabolic effects associated with OC use. A novel progestin, norgestimate (NGM) in a dose of 250 f-Lg, has been combined with 35 f-Lg of ethinyl estradiol (EE) in a new ~C, NGM/EE (Ortho-Cyclen). A number of phase II and phase III clinical studies have been conducted in the United States and Europe to assess the efficacy, safety, and tolerability of this formulation. This article summarizes the results of those studies.
Clinical studies of a norgestimate-containing OC
1197
Efficacy In the largest clinical trial conducted to date, 59,701 women were studied over 342,345 cycles (data on file at R.W. Johnson Pharmaceutical Research Institute, Raritan, N.J.). Statistical analysis indicated a use-efficacy Pearl index of 0.25 for NGM/EE.ln double-blind, randomized clinical trials comparing NGM/EE and a reference preparation, Lo Ovral, containing norgestrel, 300 f-Lg, and ethinyl estradiol, 30 f-Lg (NG/EE), there were no statistically significant differences in use or theoretic efficacy as determined by both Pearl and lifetable analyses. Safety Lipid/lipoprotein changes. NGM/EE was compared with NG/EE in two double-blind multicenter, randomized clinical trials. Combined results have been reported previously.' Subjects totaled 1473 women randomly assigned to receive NGM/EE (736 women) or NG/EE (737 women) for 24 treatment cycles. Each cycle entailed a course of 21 days on treatment followed by 7 days off medication. Differences between the two study groups were seen in the effects on lipid metabolism as assessed in fresh subjects, women who had not used an OC for 3 months before initiation of the study. NGM/EE elevated HDL cholesterol serum levels in fresh subjects. Mean levels rose steadily, beginning with treatment cycle 3 (4.9% over baseline) and continuing through the twentyfourth cycle (to a high of 10.8%). The increase was statistically significant at each cycle. In contrast, the subjects using NG/EE experienced a fall in HDL cholesterol levels over the study, ranging from a low of -10.0% at the third treatment cycle to -7.8% at the twenty-fourth cycle. These changes were also statistically significant. Statistically significant changes of even greater magnitude were noted when all subjects (versus fresh subjects only) were included in the analysis. NGM/EE proved to have a much lesser effect on LDL cholesterol serum levels in fresh subjects than did NG/EE. Although both patient groups showed a similar rise in LDL cholesterol at the third treatment cycle, these values began falling in subjects receiving NGM/EE. After cycle 24, LDL cholesterol levels were 4.3% above baseline, a statistically nonsignificant change. In contrast, LDL cholesterol levels continued to rise in subjects receiving NG/EE, reaching a high of 15.6% above baseline by cycle 24, a statistically significant change. The LDLlHDL ratio is a useful predictor of cardiovascular risk. The higher the ratio, the greater the risk. Lowering of the LDLlHDL ratio by NGM/EE reflects the dramatic increase in HDL cholesterol and minimal effect on LDL cholesterol caused by this OC (Fig. 1). In contrast, NG/EE caused a marked rise in this ratio
1198
Kafrissen
October 1992 Am J Obstel Gynecol
28.0 Q)
24.0
oS iii 20.0 en a1
CO
E
,g
o
NGM/EE NG/EE
_
18.3*
16.0 12.0
Q)
0>
8.0
c a1 s::
4.0
()
;R 0
0 -4.0
3
6
12
24
Cycle
*Statistically significant change
Fig. 1. Change in mean LDLlHDL ratio (fresh subjects). (Data reported in part from Chapdelaine A, Desmarais]-L, Derman RJ. Int] Ferdl 1989;34:347.)
6.0 oS
iii en
2.0
CO
0
a1
E
,g Q)
-4.0
()
-6.0
a1
s::
*
NGM/EE NG/EE
-2.0
c
0>
0
4.2
4.0
Q)
-8.0 -10.0
3
12
6
24
Cycle
Fig. 2. Change in mean triglyceride values with NGM/EE versus NG/EE as computed for all subjects. (Data reported in part from Chapdelaine A, Desmarais J-L, Derman RJ. lnt J Ferti! 1989;34:347.)
3
ng/mL
2
r-'--
--
O~--~~I~--~~~--~~~-
Control cycle
Cycle 3
Cycle 6
(n=19)
(n=16)
(n=12)
Fig. 3. Fibrinopeptidc A concentration was unchanged over 6 months of NGM/EE treatment in the phase II coagulation study.
of 18.3% by treatment cycle 3 and 26.6% by the twentyfourth cycle. Differences were statistically significant at all evaluation points. Estrogen increases serum triglyceride levels. Androgenic progestins such as NG generally reverse this increase, whereas selective, low androgenic progestins (e.g., NGM) have less effect. The NG-containing OC, NG/EE, reversed the estrogen-induced increase in serum triglyceride levels, causing an initial reduction from baseline of - 4.9% at treatment cycle 3, falling to a low of - 8.9% by cycle 12. In contrast, the NGMcontaining OC, NGM/EE, caused a slight rise in serum triglyceride levels, which gradually returned to baseline by treatment cycle 12 (Fig. 2). Several Phase II studies were conducted to determine the effects of NGM/EE
Clinical studies of a norgestimate-containing OC
Volume 167 l\umber 4, Part 2
140 120
I'""
-
T
T
80 -
T
1199
T
100
0/0
60
f0-
40
r-
20
-
o
Pre-cycle (n=19)
Cycle 1 (n=18)
Cycle 3 (n=16)
Cycle 6 (n=12)
Fig. 4. Antithrombin III levels remained unchanged. (From Becker H. Acta Obstet Gynecol Scand Suppl 1990; 152:33.)
40 30 Dinor (pmol/mmol)
374
381
Cycle 3
Cycle 6
29.8
20 10 0 Baseline
Fig. 5. Levels of PCI, increased with NCM/EE use.
on biochemical parameters, including carbohydrate metabolism and coagulation factors. Carbohydrate metabolism. A decrease in glucose tolerance has been observed in women taking oes. The effect of NGM/EE on carbohydrate metabolism was evaluated by Kaiser over six cycles and during followup among 21 healthy, ovulating women who had never used oes or had discontinued OCs for at least 3 months before the study." There were no significant changes in basal fasting glucose values. After oral glucose loading, slight increases in serum glucose levels occurred at I hour. At the 2 hour interval, levels had returned to baseline. Hemoglobin AI, values, a reflection of longterm serum glucose concentrations, were unchanged by NGM/EE. There were no significant changes in basal insulin values during treatment. A slight increase during cycle 3 was followed by a smaller increase at the sixth treatment cycle and a return to pretherapy levels during follow-up. These changes in insulin resistance appear to be modest relative to those reported for other oes.
However, clarification of the effect awaits randomized trials. Coagulation. The risk of thromboembolism in OC users is in part related to the dose of estrogen. Factor VII activity, which promotes coagulation, is higher in users of oes as a dose-related effect of estrogen." H Increases in fibrinogen, another coagulation enhancer, have also been linked to estrogen dose, as have modest decreases in antithrombin III, a coagulation inhibitor. H• 9 The effect of NGM/EE on coagulation was evaluated by Anger" over six cycles in 19 healthy young women who had not used oes for the previous 3 months. The concentration of fibrinopeptide A, a very sensitive marker of coagulation system activation, was unchanged during treatment with NGM/EE (Fig. 3). The anticoagulant factors antithrombin III (Fig. 4) and protein C remained stable throughout the study. The coagulation-promoting factors fibrinogen, factor VII, and factor VIII also remained essentially unchanged. The mechanisms that account for an increased risk
1200 Kafrissen
October 1992 Am J Obstet Gynecol
4.0
7
3.5
6
3.0
.---
5
ng/dl
4
3
2.5
*
ngl100mL
~
2.0
--
~
1.5
2
**
1.0 0.5
o
Control cycle
Cycle 3
Cycle 6
Follow-up cycle 2
Control cycle
Cycle 3
Cycle Follow-up 6 cycle 2
Fig. 6. SHBG levels in phase II study. (Data from Becker H. Acta Obstet Gynecol Scand Suppl 1990; 152:33.)
Fig. 7. Free testosterone levels. (Data from Becker H. Acta Obstet Gynecol Scand Suppl 1990; 152:33.)
of occlusive vascular disease in OC users have yet to be clarified. Among potential mechanisms are the following: a decrease in serum HDL cholesterol and a reduction in the level of the vasodilative and antiaggregatory prostaglandin prostacyclin (PGI.), along with a rise in levels of its antagonist, thromboxane A. (TXA.). In a study of 20 healthy women given NGM/EE through six cycles, the production of PGI. was evaluated by determinations of two urinary F. prostaglandins: dinor (2,3 dinor-6-keto-prostaglandin F.a) and 6keto (6-keto-prostaglandin F.a ) (data on file at the R. W. Johnson Pharmaceutical Research Institute). HDL (including HDL.) cholesterol and PGI. levels increased significantly in this study (p < 0.01), suggesting that HDL-cholesterol regulation of PGI. is not affected by the norge stirn ate-containing OC (Fig. 5). The study also confirmed that NGM did not affect thromboxane B. (TXB.) levels or the capacity of platelets to produce TXB., urinary excretion of which is thought to reflect platelet-derived TXA. synthesis. Thus although HDL. cholesterol may be associated with PGI. regulation, it had no effect on TXA. synthesis. The increase in PGI. and HDL cholesterol during OC use may be beneficial as it relates to the reduction of a cardiovascular risk factor.
as weight gain. Androgenic progestins will inhibit estrogen-induced elevations of SHBG. The NGM/EE formulation's lack of androgenicity was documented in Kaiser's study.6 NGM/EE caused a continuous increase in SHBG during treatment, with a peak value at cycle 6 three times higher than controls (Fig. 6). As expected, the absolute concentration of free testosterone declined, and at cycles 3 and 6, it was significantly lower than pretherapy values (Fig. 7). In this same study, prolactin levels remained unchanged from baseline, indicating that NGM/EE is a hormonally well-balanced preparation. An elevated prolactin level may reflect estrogen dominance. In a comparison study on the endocrine system, 40 healthy, ovulating women 18 to 40 years old were randomly assigned to receive NGM/EE or NG/EE. 10 Treatment consisted of a pretherapy control cycle, four cycles of therapy, and one posttherapy cycle. Comparison of SHBG levels revealed distinct differences. Although pretherapy measurements were comparable in the two cohorts, they diverged greatly once OC use began. By the fourth cycle, mean SHBG levels had more than doubled in the NGM/EE group while increasing only about 10% among NG/EE users. The difference from pretherapy values was highly statistically significant for NGM/EE (p = 0.0007). These findings support the results of the combined phase III studies in 1473 women demonstrating the low clinical androgenicity of NGM/EE. Consistent with these results, reports from the large clinical study of NGM/EE in 59,701 women continue
Endocrine function Sex hormone-binding globulin (SHBG) levels are an indicator of a drug'S androgenic potential and are inversely related to free testosterone levels. Elevations of free testosterone are associated with side effects such
Clinical studies of a norgestimate-containing OC
Volume 167 Number 4, Part 2
1201
50 DNGMlEE _ NG/EE
%
25
11,3
10.6
0 ........- 1-6
7-12
13-24
Cycles Fig. 8. Breakthrough bleeding and spotting for all cycles in the comparative study of NGM/EE and NG/EE. (Data from Corson SL. Acta Obstet Gynecol Scand Suppl 1990;152:33.)
50 D NGMlEE _NG/EE
%
25
o
2.6 2
3
2.9 4
Cycles Fig. 9. Incidence of breakthrough bleeding in comparative study.
to show a low incidence of weight gain, hirsutism, and acne. Indeed, an overall improvement in acne was seen when pre therapy versus on-therapy experience was compared.
Tolerability How consistently a formulation is used depends in large part on how well it is tolerated. Factors affecting tolerability include side effects and bleeding patterns. Although adverse effects associated with androgenicity were low in both treatment groups in the comparative study of NGM/EE and NG/EE, there was a difference for acne and weight gain. Acne occurred in 0.53% of cycles with NGM/EE versus 1.02 with NG/EE. Weight gain was the cause of discontinuation in 1.0% of subjects
taking NGM/EE (7 of 736) versus 1.4% of those taking NG/EE (10 of 737).5 Cycle control. Bleeding patterns were recorded in the comparative study of NGM/EE and NGI EE in 1473 women. 5 The incidence of breakthrough bleeding and spotting was similar for both preparations during cycles 1 through 6, 7 through 12, and 13 through 24, and differences were not significant (Fig. 8). In both groups, breakthrough bleeding and spotting followed the expected clinical course; episodes were more prevalent during the first six cycles of OC use and gradually declined thereafter in both cohorts. Examination of the first four cycles individually for breakthrough bleeding revealed no statistically significant differences, with very low rates for each group (Fig. 9). Summarization
1202 Kafrissen
October 1992 Am J Obstet Gynecol
10
DNGM/EE _ NG/EE
8 6
%
4
1.8
2 0
1-6
7-12
13-24
Cycles Fig. 10. Incidence of single amenorrheic cycles in comparative study. (Data from Corson SL. Acta Obstet Gynecol Scand SuppI1990;152:33.)
of rates of breakthrough bleeding beyond cycle 4 continued to show no change through cycle 6. Failure to experience withdrawal bleeding after a 21day pill-taking cycle was a rare event with both formulations, although there was a consistent advantage with NGM/EE (Fig. 10). By the twenty-fourth treatment cycle, 1.4% of NG/EE users had experienced at least one single amenorrheic cycle compared with 0.4% of N GM I EE users. The incidence of amenorrhea -defined as two consecutive treatment cycles without menses-was even lower (0.8% with NG/EE and 0% with NGM/EE). Duration of menses remained almost unchanged from the natural cycle during therapy. With NGM/EE, pretherapy menses lasted 5.5 days, whereas on-therapy withdrawal bleeding lasted 4.9 days. With NG/EE, pretherapy menses lasted 4.9 days, whereas on-therapy withdrawal bleeding lasted 4.2 days. Comment
Cardiovascular disease remains the leading cause of death among women in the developed world. The lifetime metabolic profile, in particular the lipid profile, is a major contributor to lifetime risk. Prescribing of OC formulations that do not worsen that profile has come to represent the prudent choice in hormone use. Several clinical studies have demonstrated NGM's lack of androgenicity and concomitant low incidence of side effects in a highly efficacious OC having excellent cycle control. An improved LDLlHDL ratio resulting from
an increased HDL and minimal effect on LDL attests to its benign influence on the lipid profile: NGM does not negate the favorable effects of estrogen. NGM appears to be a valuable addition to the hormonal armamentarium.
REFERENCES 1. LaRosa JC. Effects of oral contraceptives on circulating lipids and lipoproteins: maximizing benefit, minimizing risk. Int J Fertil 1989;34(suppl):71-84. 2. Crook D, Godsland IF, Wynn V. Oral contraceptives and coronary heart disease: modulation of glucose tolerance and plasma lipid risk factors by progestins. AM J OBSTET GVNECOL 1988;158:1612-20. 3. Perlman]A, Russel-Briefel R, Ezzati T, et al. Oral glucose tolerance and the potency of contraceptive progestins. J Chronic Dis 1985;38:857-64. 4. Stubblefield PG. Cardiovascular effects of oral contraceptives: a review. Int J Ferti! 1989;34(suppl):40-9. 5. Corson SL. Efficacy and clinical profile of a new oral contraceptive containing norgestimate. U.S. clinical trials. Acta Obstet Gynecol Scand Suppl 1990; 152:25. 6. Becker H. Supportive European data on a new oral contraceptive containing norgestimate. Acta Obstet Gynecol Scand Suppl 1990;152:33. 7. Poller L, ThompsonJM. Clotting factors during oral contraception: further report. Br Med] 1966;2:23. 8. Meade TW, Chakrabarti R, Haines AP, et al. Haemostatic, lipid and blood-pressure profiles of women on oral contraceptives containing 50 fLg or 30 fLg oestrogen. Lancet 1977;2:948. 9. Meade TW. Risks and mechanisms of cardiovascular events in users of oral contraceptives. AM] OBSTET GvNECOL 1988;158:1546. 10. Chapdelaine A, Desmarais J-L, Derman RJ. Clinical evidence of the minimal androgenic activity of norgestimate. Int] FertiI1989;34:347.
to be a selective progestin on the basis of reduced androgenicity. REFERENCES l. Wahl P, Walden C, Knopp R, et al. Effect of estrogen/progestin potency on lipid/lipoprotein cholesterol. N Engl] Med 1983;308:862-7. 2. Gaspard U]. Metabolic effects of oral contraceptives. AM ] OBSTET GYNECOL 1987;157:1029-4l. 3. Fotherby K. Oral contraceptives, lipids and cardiovascular disease. Contraception 1985;31 :367 -94. 4. Wynn V, Niththyananthan R. The effect of progestins in combined oral contraceptives on serum lipids with special reference to high-density lipoproteins. AM] OBSTET GyNECOL 1982:142:766-72. 5. Chapdelaine A, Desmarais ]-L, Derman RJ. Clinical evidence of the minimal androgenic activity of norgestimate. Int] Fertil 1989;34:347-52. 6. McGuire ]L, Phillips A, Hahn DW, et al. Pharmacologic and pharmacokinetic characteristics of norgestimate and its metabolites. AM] OBSTET GYNECOL 1990; 163:2127-3l.
October 1992 Am J Obstet Gynecol
7. Phillips A, Demarest K, Hahn DW, et al. Progestational and androgenic receptor binding affinities and in-vivo activities of norgestimate and other progestins. Contraception 1990;41 :399-410. 8. Phillips A, Hahn DW, Klimek S, McGuire]L. A comparison of the potencies and activities of progestogens used in oral contraceptives. Contraception 1987;36:181-92. 9. Phillips A. The selectivity of a new progestin. Acta Obstet Gynecol Scand 1990;69:(suppl 152):21-4. 10. Killinger ], Hahn DW, Phillips A, et al. The affinity of norgestimate for uterine progestogen receptors and its direct action on the uterus. Contraception 1985;32: 31l. Il. Phillips A, Hahn DW, McGuire]L. Relative binding affinity of norgestimate and other progestins for human sex hormone-binding globulin. Steroids 1990;55:373-5. 12. EI Makhzangy MN, Wynn V, Lawrence DM. Sex hormone binding globulin capacity as an index of oestrogenicity or androgenicity in women on oral contraceptive steroids. Clin Endocrinol 1979; 10:39-45. 13. Victor A, Weiner E, Johansson EDB. Relation between sex hormone binding globulin and d-norgestrel levels in plasma. Acta Endocrinol 1977;86:430-6.
A norgestimate-containing oral contraceptive: Review of clinical studies Michael Edwin Kafrissen, MD, MSPH Raritan, New Jersey Despite the well-documented efficacy and safety of low-dose oral contraceptives, the development of newer formulations containing highly selective progestins with minimal or no androgenic activity has been a goal of pharmaceutical research. The efficacy and safety of norgestimate, a progestin with inherently low androgenicity, in combination with ethinyl estradiol, has been examined in several phase II and phase III clinical studies, and these are reviewed. Norgestimate/ethinyl estradiol has proved to be a low-dose oral contraceptive with high selectivity that provides the cycle control of older oral contraceptive formulations with comparable efficacy. Results of comparison studies between norgestimate/ethinyl estradiol and formulations containing norgestrel, a progestin with relatively greater androgenic activity, in combination with ethinyl estradiol, are reported for effects on lipid and lipoprotein levels and carbohydrate metabolism. Norgestimate/ethinyl estradiol consistently produced statistically significant increases in high-density lipoprotein cholesterol and improvement in the ratio of low-density lipoprotein cholesterol to high-density lipoprotein. In contrast, norgestrel/ethinyl estradiol produced statistically significant decreases in high-density lipoprotein cholesterol and potentially adverse changes in the low-density/high-density lipoprotein ratio. Phase II studies have confirmed that norgestimate/ethinyl estradiol has low androgenic activity and causes minimal effect on coagulation factors and carbohydrate metabolism. (AM J OSSTET GVNECOL
1992;167:1196-202.)
Key words: Cilest, Ortho-Cyclen, low-dose oral contraceptives, norgestimate, low androgenicity
From The R. W. Johnson Pharmaceutical Research Institute. Reprint requests: Michael Edwin Kafrissen, MD, MSPH, Senior Director, Clinical Reproductive Research, R. W. Johnson Pharmaceutical Research Institute, Route 202 South, Raritan, NJ 088690602. 6/0/33746
1196
With reversible, safe, and effective birth control as the objective, oral contraceptives (OCs) have been studied more widely than any other pharmacologic agents. An important step in their development was the intro-
Volume 167 "'umber 4, Part 2
duction of low-dose, fixed estrogen/progestin combinations, which dramatically reduced the incidence of adverse reactions associated with early preparations. Lowering the estrogen content to one fifth of the original dose sharply decreased the risk of thromboembolic events; lowering the progestin content resulted in substantive improvement in metabolic profiles. Currently, in healthy women under 30 years of age and in nonsmokers up to menopause, OC use provides effective birth control associated with few problems compared with those of an unintended pregnancy and affords certain health benefits as well. Recent efforts have focused on fine-tuning OCs to minimize their contribution to risk factors associated with the development of cardiovascular disease. In early studies of OCs, it became evident that the various progestins used may have androgenic and estrogenic activity as well as progestational activity. It is the effects of progestins on lipoprotein metabolism that are most readily linked to possible increased long-term cardiovascular risk with OC use: progestins tend to increase low-density lipoprotein (LDL) cholesterol and decrease high-density lipoprotein (HDL) cholesterol, including the cardioprotective HDL2 • More pronounced effects are seen with progestins that have greater androgenic activity. 1.2 Their adverse effects are most pronounced on HDL, HDL2 , and apoprotein A-I. Carbohydrate metabolism is also sensitive to the androgenicity of OC progestins. Fixed-dose, combined OCs may increase insulin resistance, resulting in elevated plasma glucose and plasma insulin levels in response to a glucose load depending on the potency of the progestin and the dose given.' Insulin resistance is another potential contributor to long-term cardiovascular risk. In general, low-dose OCs appear to affect coagulation minimally; anticoagulant and procoagulant effects have been noted. Other factors have substantial effects on coagulation, such as smoking, obesity, family history, and blood type.' However, current use of OCs remains a risk factor for acute-event cardiovascular risk. The prudent choice of OC remains a low-dose, low androgenicity formulation that is compatible with.good safety, efficacy, and tolerability. Development of a selective progestin with a targeted response and minimal androgenicity would be a step forward in efforts to eliminate the adverse metabolic effects associated with OC use. A novel progestin, norgestimate (NGM) in a dose of 250 f-Lg, has been combined with 35 f-Lg of ethinyl estradiol (EE) in a new ~C, NGM/EE (Ortho-Cyclen). A number of phase II and phase III clinical studies have been conducted in the United States and Europe to assess the efficacy, safety, and tolerability of this formulation. This article summarizes the results of those studies.
Clinical studies of a norgestimate-containing OC
1197
Efficacy In the largest clinical trial conducted to date, 59,701 women were studied over 342,345 cycles (data on file at R.W. Johnson Pharmaceutical Research Institute, Raritan, N.J.). Statistical analysis indicated a use-efficacy Pearl index of 0.25 for NGM/EE.ln double-blind, randomized clinical trials comparing NGM/EE and a reference preparation, Lo Ovral, containing norgestrel, 300 f-Lg, and ethinyl estradiol, 30 f-Lg (NG/EE), there were no statistically significant differences in use or theoretic efficacy as determined by both Pearl and lifetable analyses. Safety Lipid/lipoprotein changes. NGM/EE was compared with NG/EE in two double-blind multicenter, randomized clinical trials. Combined results have been reported previously.' Subjects totaled 1473 women randomly assigned to receive NGM/EE (736 women) or NG/EE (737 women) for 24 treatment cycles. Each cycle entailed a course of 21 days on treatment followed by 7 days off medication. Differences between the two study groups were seen in the effects on lipid metabolism as assessed in fresh subjects, women who had not used an OC for 3 months before initiation of the study. NGM/EE elevated HDL cholesterol serum levels in fresh subjects. Mean levels rose steadily, beginning with treatment cycle 3 (4.9% over baseline) and continuing through the twentyfourth cycle (to a high of 10.8%). The increase was statistically significant at each cycle. In contrast, the subjects using NG/EE experienced a fall in HDL cholesterol levels over the study, ranging from a low of -10.0% at the third treatment cycle to -7.8% at the twenty-fourth cycle. These changes were also statistically significant. Statistically significant changes of even greater magnitude were noted when all subjects (versus fresh subjects only) were included in the analysis. NGM/EE proved to have a much lesser effect on LDL cholesterol serum levels in fresh subjects than did NG/EE. Although both patient groups showed a similar rise in LDL cholesterol at the third treatment cycle, these values began falling in subjects receiving NGM/EE. After cycle 24, LDL cholesterol levels were 4.3% above baseline, a statistically nonsignificant change. In contrast, LDL cholesterol levels continued to rise in subjects receiving NG/EE, reaching a high of 15.6% above baseline by cycle 24, a statistically significant change. The LDLlHDL ratio is a useful predictor of cardiovascular risk. The higher the ratio, the greater the risk. Lowering of the LDLlHDL ratio by NGM/EE reflects the dramatic increase in HDL cholesterol and minimal effect on LDL cholesterol caused by this OC (Fig. 1). In contrast, NG/EE caused a marked rise in this ratio
1198
Kafrissen
October 1992 Am J Obstel Gynecol
28.0 Q)
24.0
oS iii 20.0 en a1
CO
E
,g
o
NGM/EE NG/EE
_
18.3*
16.0 12.0
Q)
0>
8.0
c a1 s::
4.0
()
;R 0
0 -4.0
3
6
12
24
Cycle
*Statistically significant change
Fig. 1. Change in mean LDLlHDL ratio (fresh subjects). (Data reported in part from Chapdelaine A, Desmarais]-L, Derman RJ. Int] Ferdl 1989;34:347.)
6.0 oS
iii en
2.0
CO
0
a1
E
,g Q)
-4.0
()
-6.0
a1
s::
*
NGM/EE NG/EE
-2.0
c
0>
0
4.2
4.0
Q)
-8.0 -10.0
3
12
6
24
Cycle
Fig. 2. Change in mean triglyceride values with NGM/EE versus NG/EE as computed for all subjects. (Data reported in part from Chapdelaine A, Desmarais J-L, Derman RJ. lnt J Ferti! 1989;34:347.)
3
ng/mL
2
r-'--
--
O~--~~I~--~~~--~~~-
Control cycle
Cycle 3
Cycle 6
(n=19)
(n=16)
(n=12)
Fig. 3. Fibrinopeptidc A concentration was unchanged over 6 months of NGM/EE treatment in the phase II coagulation study.
of 18.3% by treatment cycle 3 and 26.6% by the twentyfourth cycle. Differences were statistically significant at all evaluation points. Estrogen increases serum triglyceride levels. Androgenic progestins such as NG generally reverse this increase, whereas selective, low androgenic progestins (e.g., NGM) have less effect. The NG-containing OC, NG/EE, reversed the estrogen-induced increase in serum triglyceride levels, causing an initial reduction from baseline of - 4.9% at treatment cycle 3, falling to a low of - 8.9% by cycle 12. In contrast, the NGMcontaining OC, NGM/EE, caused a slight rise in serum triglyceride levels, which gradually returned to baseline by treatment cycle 12 (Fig. 2). Several Phase II studies were conducted to determine the effects of NGM/EE
Clinical studies of a norgestimate-containing OC
Volume 167 l\umber 4, Part 2
140 120
I'""
-
T
T
80 -
T
1199
T
100
0/0
60
f0-
40
r-
20
-
o
Pre-cycle (n=19)
Cycle 1 (n=18)
Cycle 3 (n=16)
Cycle 6 (n=12)
Fig. 4. Antithrombin III levels remained unchanged. (From Becker H. Acta Obstet Gynecol Scand Suppl 1990; 152:33.)
40 30 Dinor (pmol/mmol)
374
381
Cycle 3
Cycle 6
29.8
20 10 0 Baseline
Fig. 5. Levels of PCI, increased with NCM/EE use.
on biochemical parameters, including carbohydrate metabolism and coagulation factors. Carbohydrate metabolism. A decrease in glucose tolerance has been observed in women taking oes. The effect of NGM/EE on carbohydrate metabolism was evaluated by Kaiser over six cycles and during followup among 21 healthy, ovulating women who had never used oes or had discontinued OCs for at least 3 months before the study." There were no significant changes in basal fasting glucose values. After oral glucose loading, slight increases in serum glucose levels occurred at I hour. At the 2 hour interval, levels had returned to baseline. Hemoglobin AI, values, a reflection of longterm serum glucose concentrations, were unchanged by NGM/EE. There were no significant changes in basal insulin values during treatment. A slight increase during cycle 3 was followed by a smaller increase at the sixth treatment cycle and a return to pretherapy levels during follow-up. These changes in insulin resistance appear to be modest relative to those reported for other oes.
However, clarification of the effect awaits randomized trials. Coagulation. The risk of thromboembolism in OC users is in part related to the dose of estrogen. Factor VII activity, which promotes coagulation, is higher in users of oes as a dose-related effect of estrogen." H Increases in fibrinogen, another coagulation enhancer, have also been linked to estrogen dose, as have modest decreases in antithrombin III, a coagulation inhibitor. H• 9 The effect of NGM/EE on coagulation was evaluated by Anger" over six cycles in 19 healthy young women who had not used oes for the previous 3 months. The concentration of fibrinopeptide A, a very sensitive marker of coagulation system activation, was unchanged during treatment with NGM/EE (Fig. 3). The anticoagulant factors antithrombin III (Fig. 4) and protein C remained stable throughout the study. The coagulation-promoting factors fibrinogen, factor VII, and factor VIII also remained essentially unchanged. The mechanisms that account for an increased risk
1200 Kafrissen
October 1992 Am J Obstet Gynecol
4.0
7
3.5
6
3.0
.---
5
ng/dl
4
3
2.5
*
ngl100mL
~
2.0
--
~
1.5
2
**
1.0 0.5
o
Control cycle
Cycle 3
Cycle 6
Follow-up cycle 2
Control cycle
Cycle 3
Cycle Follow-up 6 cycle 2
Fig. 6. SHBG levels in phase II study. (Data from Becker H. Acta Obstet Gynecol Scand Suppl 1990; 152:33.)
Fig. 7. Free testosterone levels. (Data from Becker H. Acta Obstet Gynecol Scand Suppl 1990; 152:33.)
of occlusive vascular disease in OC users have yet to be clarified. Among potential mechanisms are the following: a decrease in serum HDL cholesterol and a reduction in the level of the vasodilative and antiaggregatory prostaglandin prostacyclin (PGI.), along with a rise in levels of its antagonist, thromboxane A. (TXA.). In a study of 20 healthy women given NGM/EE through six cycles, the production of PGI. was evaluated by determinations of two urinary F. prostaglandins: dinor (2,3 dinor-6-keto-prostaglandin F.a) and 6keto (6-keto-prostaglandin F.a ) (data on file at the R. W. Johnson Pharmaceutical Research Institute). HDL (including HDL.) cholesterol and PGI. levels increased significantly in this study (p < 0.01), suggesting that HDL-cholesterol regulation of PGI. is not affected by the norge stirn ate-containing OC (Fig. 5). The study also confirmed that NGM did not affect thromboxane B. (TXB.) levels or the capacity of platelets to produce TXB., urinary excretion of which is thought to reflect platelet-derived TXA. synthesis. Thus although HDL. cholesterol may be associated with PGI. regulation, it had no effect on TXA. synthesis. The increase in PGI. and HDL cholesterol during OC use may be beneficial as it relates to the reduction of a cardiovascular risk factor.
as weight gain. Androgenic progestins will inhibit estrogen-induced elevations of SHBG. The NGM/EE formulation's lack of androgenicity was documented in Kaiser's study.6 NGM/EE caused a continuous increase in SHBG during treatment, with a peak value at cycle 6 three times higher than controls (Fig. 6). As expected, the absolute concentration of free testosterone declined, and at cycles 3 and 6, it was significantly lower than pretherapy values (Fig. 7). In this same study, prolactin levels remained unchanged from baseline, indicating that NGM/EE is a hormonally well-balanced preparation. An elevated prolactin level may reflect estrogen dominance. In a comparison study on the endocrine system, 40 healthy, ovulating women 18 to 40 years old were randomly assigned to receive NGM/EE or NG/EE. 10 Treatment consisted of a pretherapy control cycle, four cycles of therapy, and one posttherapy cycle. Comparison of SHBG levels revealed distinct differences. Although pretherapy measurements were comparable in the two cohorts, they diverged greatly once OC use began. By the fourth cycle, mean SHBG levels had more than doubled in the NGM/EE group while increasing only about 10% among NG/EE users. The difference from pretherapy values was highly statistically significant for NGM/EE (p = 0.0007). These findings support the results of the combined phase III studies in 1473 women demonstrating the low clinical androgenicity of NGM/EE. Consistent with these results, reports from the large clinical study of NGM/EE in 59,701 women continue
Endocrine function Sex hormone-binding globulin (SHBG) levels are an indicator of a drug'S androgenic potential and are inversely related to free testosterone levels. Elevations of free testosterone are associated with side effects such
Clinical studies of a norgestimate-containing OC
Volume 167 Number 4, Part 2
1201
50 DNGMlEE _ NG/EE
%
25
11,3
10.6
0 ........- 1-6
7-12
13-24
Cycles Fig. 8. Breakthrough bleeding and spotting for all cycles in the comparative study of NGM/EE and NG/EE. (Data from Corson SL. Acta Obstet Gynecol Scand Suppl 1990;152:33.)
50 D NGMlEE _NG/EE
%
25
o
2.6 2
3
2.9 4
Cycles Fig. 9. Incidence of breakthrough bleeding in comparative study.
to show a low incidence of weight gain, hirsutism, and acne. Indeed, an overall improvement in acne was seen when pre therapy versus on-therapy experience was compared.
Tolerability How consistently a formulation is used depends in large part on how well it is tolerated. Factors affecting tolerability include side effects and bleeding patterns. Although adverse effects associated with androgenicity were low in both treatment groups in the comparative study of NGM/EE and NG/EE, there was a difference for acne and weight gain. Acne occurred in 0.53% of cycles with NGM/EE versus 1.02 with NG/EE. Weight gain was the cause of discontinuation in 1.0% of subjects
taking NGM/EE (7 of 736) versus 1.4% of those taking NG/EE (10 of 737).5 Cycle control. Bleeding patterns were recorded in the comparative study of NGM/EE and NGI EE in 1473 women. 5 The incidence of breakthrough bleeding and spotting was similar for both preparations during cycles 1 through 6, 7 through 12, and 13 through 24, and differences were not significant (Fig. 8). In both groups, breakthrough bleeding and spotting followed the expected clinical course; episodes were more prevalent during the first six cycles of OC use and gradually declined thereafter in both cohorts. Examination of the first four cycles individually for breakthrough bleeding revealed no statistically significant differences, with very low rates for each group (Fig. 9). Summarization
1202 Kafrissen
October 1992 Am J Obstet Gynecol
10
DNGM/EE _ NG/EE
8 6
%
4
1.8
2 0
1-6
7-12
13-24
Cycles Fig. 10. Incidence of single amenorrheic cycles in comparative study. (Data from Corson SL. Acta Obstet Gynecol Scand SuppI1990;152:33.)
of rates of breakthrough bleeding beyond cycle 4 continued to show no change through cycle 6. Failure to experience withdrawal bleeding after a 21day pill-taking cycle was a rare event with both formulations, although there was a consistent advantage with NGM/EE (Fig. 10). By the twenty-fourth treatment cycle, 1.4% of NG/EE users had experienced at least one single amenorrheic cycle compared with 0.4% of N GM I EE users. The incidence of amenorrhea -defined as two consecutive treatment cycles without menses-was even lower (0.8% with NG/EE and 0% with NGM/EE). Duration of menses remained almost unchanged from the natural cycle during therapy. With NGM/EE, pretherapy menses lasted 5.5 days, whereas on-therapy withdrawal bleeding lasted 4.9 days. With NG/EE, pretherapy menses lasted 4.9 days, whereas on-therapy withdrawal bleeding lasted 4.2 days. Comment
Cardiovascular disease remains the leading cause of death among women in the developed world. The lifetime metabolic profile, in particular the lipid profile, is a major contributor to lifetime risk. Prescribing of OC formulations that do not worsen that profile has come to represent the prudent choice in hormone use. Several clinical studies have demonstrated NGM's lack of androgenicity and concomitant low incidence of side effects in a highly efficacious OC having excellent cycle control. An improved LDLlHDL ratio resulting from
an increased HDL and minimal effect on LDL attests to its benign influence on the lipid profile: NGM does not negate the favorable effects of estrogen. NGM appears to be a valuable addition to the hormonal armamentarium.
REFERENCES 1. LaRosa JC. Effects of oral contraceptives on circulating lipids and lipoproteins: maximizing benefit, minimizing risk. Int J Fertil 1989;34(suppl):71-84. 2. Crook D, Godsland IF, Wynn V. Oral contraceptives and coronary heart disease: modulation of glucose tolerance and plasma lipid risk factors by progestins. AM J OBSTET GVNECOL 1988;158:1612-20. 3. Perlman]A, Russel-Briefel R, Ezzati T, et al. Oral glucose tolerance and the potency of contraceptive progestins. J Chronic Dis 1985;38:857-64. 4. Stubblefield PG. Cardiovascular effects of oral contraceptives: a review. Int J Ferti! 1989;34(suppl):40-9. 5. Corson SL. Efficacy and clinical profile of a new oral contraceptive containing norgestimate. U.S. clinical trials. Acta Obstet Gynecol Scand Suppl 1990; 152:25. 6. Becker H. Supportive European data on a new oral contraceptive containing norgestimate. Acta Obstet Gynecol Scand Suppl 1990;152:33. 7. Poller L, ThompsonJM. Clotting factors during oral contraception: further report. Br Med] 1966;2:23. 8. Meade TW, Chakrabarti R, Haines AP, et al. Haemostatic, lipid and blood-pressure profiles of women on oral contraceptives containing 50 fLg or 30 fLg oestrogen. Lancet 1977;2:948. 9. Meade TW. Risks and mechanisms of cardiovascular events in users of oral contraceptives. AM] OBSTET GvNECOL 1988;158:1546. 10. Chapdelaine A, Desmarais J-L, Derman RJ. Clinical evidence of the minimal androgenic activity of norgestimate. Int] FertiI1989;34:347.
