FERTILITY AND STERILITY Copyright
©
Vol. 57, No. 1, January 1992
1992 The American Fertility Society
Printed on acid-free paper in U.S.A.
Gonadotropin suppression for the treatment of karyotypically normal spontaneous premature ovarian failure: a controlled trial*
Lawrence M. Nelson, M.D.t:j: Lorene M. Kimzey, R.N., B.S.§ Beverly J. White, M.D. II George R. Merriam, M.D.:j: National Institutes of Health, Bethesda, Maryland
Objective: To determine if gonadotropin suppression improves ovarian follicle function or ovulation rates in patients with karyotypically normal spontaneous premature ovarian failure. Design: Prospective, double-blind, placebo-controlled, crossover trial. Setting: Tertiary care research institution. Interventions: Two intervention phases lasting 4 months each: one placebo phase, and one treatment phase during which each patient received daily subcutaneous injections of 300 JLg of the gonadotropin-releasing hormone agonist (GnRH-a) deslorelin. During both phases, patients took a standardized estrogen (E) replacement regimen. Patients, Participants: Twenty-six patients with karyotypically normal spontaneous premature ovarian failure ranging in age from 18 to 39 years. Main Outcome Measures: We measured serum estradiol (E 2 ) and progesterone (P) levels weekly during the 2 months after each intervention. We defined a serum E 2 > 50 pg/mL (184 pmol/L) as evidence for ovarian follicle function and a serum P > 3.0 ng/mL (9.5 nmol/L) as evidence for ovulation. Results: The GnRH -a therapy did not significantly enhance recovery of ovarian follicle function or the chance of ovulation. The power to detect a 40% and a 33% ovulation success rate with therapy was 0.95 and 0.83, respectively. We found evidence for ovarian follicle function in 11 of 23 women (48%), and 4 women (17%) ovulated. Conclusions: Patients with karyotypically normal spontaneous premature ovarian failure treated with E replacement did not benefit from the additional gonadotropin suppression achieved with GnRH-a. Because these patients have a significant possibility of spontaneous remission, attempts to induce ovulation should be limited to controlled trials designed to determine safety and effectiveness. Fertil Steril 1992;57:50-5 Key Words: Hypergonadotropic amenorrhea, premature ovarian failure
No therapy to induce ovulation for patients with premature ovarian failure has been proven effective
Received July 1, 1991; revised and accepted September 5, 1991. * Presented at the 73rd Annual Meeting of The Endocrine Society, Washington, D.C., June 19 to 22, 1991. t Reprint requests: Lawrence M. Nelson, M.D., Developmental Endocrinology Branch, National Institute of Child Health and Human Development, National Institutes of Health, Building 10, Room 10N262, Bethesda, Maryland 20892. t Developmental Endocrinology Branch, National Institute of Child Health and Human Development.
50
Nelson et al.
Premature ovarian failure
by prospective controlled study. Hertz (1) observed in 1960 that some patients with ovarian failure have histologically normal appearing primordial follicles, and several reports have demonstrated the presence of normal primordial follicles in patients with autoimmune premature ovarian failure (2-6). At one time it was suggested that a single follicle-stimulat-
§ Nursing Department, Clinical Center. II Laboratory of Chemical Biology, National Institute of Diabetes, and Digestive and Kidney Diseases.
Fertility and Sterility
ing hormone (FSH) value > 40 IU/L might indicate irreversible ovarian failure (7). This view was challenged by a report of intermittent ovarian failure (8), and by numerous reports of pregnancy occurring after the established diagnosis of premature ovarian failure (9-11). Nevertheless, remission of premature ovarian failure was considered a rare event until Rebar et al. (12) detected evidence of functioning ovarian follicles, as demonstrated by serial blood estradiol (E 2) determinations, in 9 of 18 patients with premature ovarian failure (defined by a single FSH value> 40 IU/L). The presence of associated autoimmune disorders in patients with premature ovarian failure suggests that up to 40% may have their ovarian failure mediated by an immune process (13, 14). Indeed, these patients as a group have evidence for increased T cell activation (15, 16). However, it is difficult to select patients with autoimmune mediated ovarian failure. The accuracy with which antiovarian antibody tests detect autoimmune ovarian failure is unknown (4). Moreover, in a portion of patients the disease may be T cell mediated rather than antibody mediated (17). Cases have been reported in which a cellular autoimmune process selectively attacks only developing follicles and spares primordial follicles (2, 4-6). These patients should be particularly good candidates for remission if specific therapy could be devised to selectively reduce the immune process disturbing normal ovarian follicle function. Empirical therapies are frequently used despite a lack of proven benefit. These therapies have included attempts to increase ovarian stimulation with exogenous gonadotropins or to suppress gonadotropins with high-dose estrogen (E) or gonadotropin-releasing hormone analog (GnRH-a). Gonadotropin stimulation of developing follicles may be a factor in maintaining the production of antigens that incite the autoimmune ovarian attack. In some case reports, the intensity of the cellular infiltrate has been positively correlated with the degree of ovarian follicle maturation (2, 4-6); this suggests the putative antigen may be gonadotropin-dependent. We therefore studied the effects of potent gonadotropin suppression with an agonist analogue of GnRH as a means to temporarily reduce the ovarian antigenic load, and by this means, induce transient remission in patients with premature ovarian failure. A role for trophic hormone suppression in another autoimmune endocrine disorder, Graves' disease, was suggested in 1967 (18), and pituitary thyrotropin suppression has recently been reported to maintain remission in that disorder (19). Vol. 57, No. 1, January 1992
MATERIALS AND METHODS Recruitment
We recruited patients with spontaneous premature ovarian failure by letters to physicians and notices in medical journals. To qualify for referral, patients were to be 40 IU /Lon at least two occasions, at least 1 month apart. Patients with a contraindication to E therapy or pregnancy were not candidates. Patients with known abnormal karyotypes or evidence to suggest iatrogenic ovarian failure were also not candidates for referral. Karyotypes
We obtained karyotypes on all patients. Peripheral blood samples were cultured 72 to 96 hours in RPMI -1640 medium with standard supplements and mitogen. A minimum of 30 Giemsa-trypsin banded metaphases at the 400 to 550 band level of resolution were counted and three karyotypes prepared for each subject; 20 additional metaphases were screened if any X chromosomal abnormalities were seen. Of the first 28 patients we studied, we discovered 2 patients who were not candidates because of mosaic karyotypes 45,X/46,X,del(X)(q21.3) and 45,X/46,X,r(X). Patients
We enrolled the first 26 women referred to us with premature ovarian failure who satisfied these entry requirements. The patient group had a median age of 33 years (range 18 to 39). Ovarian failure had been diagnosed at a median age of 30 years (range 15 to 38), and the median time since diagnosis was 2 years (range 0.33 to 12). Two women (8%) had primary amenorrhea. The patients with secondary amenorrhea had a median duration of 1 year of amenorrhea preceding diagnosis (range 0.33 to 3.0 years). Thirteen women (50%) had previously been pregnant, and 10 women (38%) had previously given birth. Twenty-two of 26 (85%), had received Ereplacement therapy before their visit to the National Institutes of Health. Eight patients (31%) had received prior unsuccessful clomiphene citrate (CC) therapy, seven patients (27%) had received prior unsuccessful human menopausal gonadotropin (hMG) therapy, and one patient had had a prior successful pregnancy after therapy employing oocyte donation. Six patients (26%) had hypothyroidism, 1 patient had Addison's disease, and 1 patient had Raynaud's Nelson et al. Premature ovarian failure
51
syndrome. None had diabetes mellitus. Six patients with undiagnosed prior illnesses had the following problems: 1 patient had a severe normochromic normocytic anemia during pregnancy and required several transfusions; 2 patients had prior transient pain, swelling, and erythema of finger joints that had not been evaluated; and 3 patients had prior prolonged febrile illnesses that required hospitalization and that led to no definitive diagnosis. Nine patients (35%) had an antinuclear antibody titer ~1: 40, and 2 (8%) had a rheumatoid factor titer ~1: 320. Three patients (12%) had an antithyroglobulin antibody titer ~1:10, and 5 (22%) had an antithyroid microsomal antibody titer ~1:100. Four patients (15%) had an antiparietal cell antibody titer ~1:10, but none had pernicious anemia. None of the 26 patients had antiovarian antibodies as detected by immunohistochemistry using rabbit ovary, but the patient with Addison's disease had antisteroid cell antibodies detected by this method (20). Intervention
The intervention was prospective and administered in a randomized, double-blind, placebo-controlled, crossover design manner. The National Institute of Child Health and Human Development Clinical Research Subpanel reviewed and approved the study. Patients underwent two treatment phases of 4 months separated by a washout period of at least 2 months. During one treatment phase, each patient received daily subcutaneous (SC) injections of placebo; during the other treatment phase, each patient received daily SC injections of 300 J.Lg of the GnRH -a deslorelin (D-Trp 6 - Pro9 - NH 2 - EthylamideGnRH; Bachem Bioscience Inc., Philadelphia, PA). The Pharmaceutical Development Service, Clinical Center, National Institutes of Health (Bethesda, MD), prepared the injections for use in a blinded manner and also randomly determined the order of treatments for each patient. During both treatment phases, patients took a 2mg dose of oral micronized E 2 daily (Estrace; Mead Johnson Laboratories, Evansville, IN) and a 10-mg/d dose of medroxyprogesterone acetate from the 1st to the 14th calendar day of each month (Provera; The Upjohn Company, Kalamazoo, MI). After each 4-month treatment phase, the injections and hormonal replacement therapy were stopped, and patients were monitored for return of ovarian function. Before they began the second treatment phase, patients returned to the National Institutes of Health Clinical Center, and we reconfirmed hypergonadotropic amenorrhea. 52
Nelson et al.
Premature ovarian failure
Outcome
After each treatment phase, patients had weekly serum samples drawn for 8 weeks; the serum was stored frozen and subsequently sent to us for assay. We measured serum E 2 , progesterone (P), FSH, and luteinizing hormone (LH) by radioimmunoassay methods reported previously (21). We prospectively defined a serum E 2 > 50 pg/mL (184 pmoljL) as representing ovarian follicle function (22) and a serum P > 3 ng/mL (9.5 nmol/L) as representing ovulation (23). For statistical analysis we used Wilcoxon's two-sample test, Fisher's exact test, and X 2 as appropriate; we set P < 0.05 to be significant.
RESULTS Twenty-three women completed both phases of the study. One woman withdrew for personal reasons, one woman withdrew after she developed menometrorrhagia while taking E replacement therapy, and one woman withdrew while taking placebo injections because of the acute onset of newly diagnosed psoriatic arthritis requiring therapy. Deslorelin plus E replacement therapy suppressed serum FSH and LH levels to significantly lower levels (FSH 5.8 ± 1.6, LH 11 ± 0.9) (mean ± SEM) than placebo plus E/progestin replacement therapy (FSH 66 ± 9.6, LH 45 ± 6.7) (P < 0.0001) (Fig. 1). After finishing deslorelin and E replacement ther-
150 140 130 120 110 100
...2 .....
•
8 0
o FSH
eLH 0
©
90 80 70 80 50 40 30 20 10
0
~ 0
• ••
•
• __._
0 0
§ 8 0
0 Placebo
I
•I
0 0
•
+
Drug
Figure 1 Values of FSH and LH at the end of the placebo phase (placebo injections and E replacement therapy) and at the end of the treatment phase (GnRH-a injections combined with E replacement therapy). The horizontal lines represent the respective means for the 23 women who completed the study. The means are significantly different (P < 0.0001).
Fertility and Sterility
apy, five women (22%) demonstrated ovarian follicle function, as compared with eight women (35%) with ovarian follicle function after placebo and E replacement therapy (Fig. 2). After finishing deslorelin and E replacement therapy, three women (13%) ovulated as compared with two women (9%) who did so after placebo and E replacement therapy (Fig. 3). Two patients had ovarian follicle function after both phases, and one of these patients ovulated after both phases. Both patients had clearly demonstrated ovarian failure before beginning the second phase. Deslorelin combined with E replacement therapy did not significantly increase the proportion of patients with ovarian follicle function or ovulation despite the more profound gonadotropin suppression obtained with this therapy. Overall, 11 patients (48%) demonstrated ovarian follicle function and 4 of these patients (17%) ovulated during this study. Patients with 2 or more years since the time of their diagnosis fared no worse than patients with
©
Vol. 57, No. 1, January 1992
1992 The American Fertility Society
Printed on acid-free paper in U.S.A.
Gonadotropin suppression for the treatment of karyotypically normal spontaneous premature ovarian failure: a controlled trial*
Lawrence M. Nelson, M.D.t:j: Lorene M. Kimzey, R.N., B.S.§ Beverly J. White, M.D. II George R. Merriam, M.D.:j: National Institutes of Health, Bethesda, Maryland
Objective: To determine if gonadotropin suppression improves ovarian follicle function or ovulation rates in patients with karyotypically normal spontaneous premature ovarian failure. Design: Prospective, double-blind, placebo-controlled, crossover trial. Setting: Tertiary care research institution. Interventions: Two intervention phases lasting 4 months each: one placebo phase, and one treatment phase during which each patient received daily subcutaneous injections of 300 JLg of the gonadotropin-releasing hormone agonist (GnRH-a) deslorelin. During both phases, patients took a standardized estrogen (E) replacement regimen. Patients, Participants: Twenty-six patients with karyotypically normal spontaneous premature ovarian failure ranging in age from 18 to 39 years. Main Outcome Measures: We measured serum estradiol (E 2 ) and progesterone (P) levels weekly during the 2 months after each intervention. We defined a serum E 2 > 50 pg/mL (184 pmol/L) as evidence for ovarian follicle function and a serum P > 3.0 ng/mL (9.5 nmol/L) as evidence for ovulation. Results: The GnRH -a therapy did not significantly enhance recovery of ovarian follicle function or the chance of ovulation. The power to detect a 40% and a 33% ovulation success rate with therapy was 0.95 and 0.83, respectively. We found evidence for ovarian follicle function in 11 of 23 women (48%), and 4 women (17%) ovulated. Conclusions: Patients with karyotypically normal spontaneous premature ovarian failure treated with E replacement did not benefit from the additional gonadotropin suppression achieved with GnRH-a. Because these patients have a significant possibility of spontaneous remission, attempts to induce ovulation should be limited to controlled trials designed to determine safety and effectiveness. Fertil Steril 1992;57:50-5 Key Words: Hypergonadotropic amenorrhea, premature ovarian failure
No therapy to induce ovulation for patients with premature ovarian failure has been proven effective
Received July 1, 1991; revised and accepted September 5, 1991. * Presented at the 73rd Annual Meeting of The Endocrine Society, Washington, D.C., June 19 to 22, 1991. t Reprint requests: Lawrence M. Nelson, M.D., Developmental Endocrinology Branch, National Institute of Child Health and Human Development, National Institutes of Health, Building 10, Room 10N262, Bethesda, Maryland 20892. t Developmental Endocrinology Branch, National Institute of Child Health and Human Development.
50
Nelson et al.
Premature ovarian failure
by prospective controlled study. Hertz (1) observed in 1960 that some patients with ovarian failure have histologically normal appearing primordial follicles, and several reports have demonstrated the presence of normal primordial follicles in patients with autoimmune premature ovarian failure (2-6). At one time it was suggested that a single follicle-stimulat-
§ Nursing Department, Clinical Center. II Laboratory of Chemical Biology, National Institute of Diabetes, and Digestive and Kidney Diseases.
Fertility and Sterility
ing hormone (FSH) value > 40 IU/L might indicate irreversible ovarian failure (7). This view was challenged by a report of intermittent ovarian failure (8), and by numerous reports of pregnancy occurring after the established diagnosis of premature ovarian failure (9-11). Nevertheless, remission of premature ovarian failure was considered a rare event until Rebar et al. (12) detected evidence of functioning ovarian follicles, as demonstrated by serial blood estradiol (E 2) determinations, in 9 of 18 patients with premature ovarian failure (defined by a single FSH value> 40 IU/L). The presence of associated autoimmune disorders in patients with premature ovarian failure suggests that up to 40% may have their ovarian failure mediated by an immune process (13, 14). Indeed, these patients as a group have evidence for increased T cell activation (15, 16). However, it is difficult to select patients with autoimmune mediated ovarian failure. The accuracy with which antiovarian antibody tests detect autoimmune ovarian failure is unknown (4). Moreover, in a portion of patients the disease may be T cell mediated rather than antibody mediated (17). Cases have been reported in which a cellular autoimmune process selectively attacks only developing follicles and spares primordial follicles (2, 4-6). These patients should be particularly good candidates for remission if specific therapy could be devised to selectively reduce the immune process disturbing normal ovarian follicle function. Empirical therapies are frequently used despite a lack of proven benefit. These therapies have included attempts to increase ovarian stimulation with exogenous gonadotropins or to suppress gonadotropins with high-dose estrogen (E) or gonadotropin-releasing hormone analog (GnRH-a). Gonadotropin stimulation of developing follicles may be a factor in maintaining the production of antigens that incite the autoimmune ovarian attack. In some case reports, the intensity of the cellular infiltrate has been positively correlated with the degree of ovarian follicle maturation (2, 4-6); this suggests the putative antigen may be gonadotropin-dependent. We therefore studied the effects of potent gonadotropin suppression with an agonist analogue of GnRH as a means to temporarily reduce the ovarian antigenic load, and by this means, induce transient remission in patients with premature ovarian failure. A role for trophic hormone suppression in another autoimmune endocrine disorder, Graves' disease, was suggested in 1967 (18), and pituitary thyrotropin suppression has recently been reported to maintain remission in that disorder (19). Vol. 57, No. 1, January 1992
MATERIALS AND METHODS Recruitment
We recruited patients with spontaneous premature ovarian failure by letters to physicians and notices in medical journals. To qualify for referral, patients were to be 40 IU /Lon at least two occasions, at least 1 month apart. Patients with a contraindication to E therapy or pregnancy were not candidates. Patients with known abnormal karyotypes or evidence to suggest iatrogenic ovarian failure were also not candidates for referral. Karyotypes
We obtained karyotypes on all patients. Peripheral blood samples were cultured 72 to 96 hours in RPMI -1640 medium with standard supplements and mitogen. A minimum of 30 Giemsa-trypsin banded metaphases at the 400 to 550 band level of resolution were counted and three karyotypes prepared for each subject; 20 additional metaphases were screened if any X chromosomal abnormalities were seen. Of the first 28 patients we studied, we discovered 2 patients who were not candidates because of mosaic karyotypes 45,X/46,X,del(X)(q21.3) and 45,X/46,X,r(X). Patients
We enrolled the first 26 women referred to us with premature ovarian failure who satisfied these entry requirements. The patient group had a median age of 33 years (range 18 to 39). Ovarian failure had been diagnosed at a median age of 30 years (range 15 to 38), and the median time since diagnosis was 2 years (range 0.33 to 12). Two women (8%) had primary amenorrhea. The patients with secondary amenorrhea had a median duration of 1 year of amenorrhea preceding diagnosis (range 0.33 to 3.0 years). Thirteen women (50%) had previously been pregnant, and 10 women (38%) had previously given birth. Twenty-two of 26 (85%), had received Ereplacement therapy before their visit to the National Institutes of Health. Eight patients (31%) had received prior unsuccessful clomiphene citrate (CC) therapy, seven patients (27%) had received prior unsuccessful human menopausal gonadotropin (hMG) therapy, and one patient had had a prior successful pregnancy after therapy employing oocyte donation. Six patients (26%) had hypothyroidism, 1 patient had Addison's disease, and 1 patient had Raynaud's Nelson et al. Premature ovarian failure
51
syndrome. None had diabetes mellitus. Six patients with undiagnosed prior illnesses had the following problems: 1 patient had a severe normochromic normocytic anemia during pregnancy and required several transfusions; 2 patients had prior transient pain, swelling, and erythema of finger joints that had not been evaluated; and 3 patients had prior prolonged febrile illnesses that required hospitalization and that led to no definitive diagnosis. Nine patients (35%) had an antinuclear antibody titer ~1: 40, and 2 (8%) had a rheumatoid factor titer ~1: 320. Three patients (12%) had an antithyroglobulin antibody titer ~1:10, and 5 (22%) had an antithyroid microsomal antibody titer ~1:100. Four patients (15%) had an antiparietal cell antibody titer ~1:10, but none had pernicious anemia. None of the 26 patients had antiovarian antibodies as detected by immunohistochemistry using rabbit ovary, but the patient with Addison's disease had antisteroid cell antibodies detected by this method (20). Intervention
The intervention was prospective and administered in a randomized, double-blind, placebo-controlled, crossover design manner. The National Institute of Child Health and Human Development Clinical Research Subpanel reviewed and approved the study. Patients underwent two treatment phases of 4 months separated by a washout period of at least 2 months. During one treatment phase, each patient received daily subcutaneous (SC) injections of placebo; during the other treatment phase, each patient received daily SC injections of 300 J.Lg of the GnRH -a deslorelin (D-Trp 6 - Pro9 - NH 2 - EthylamideGnRH; Bachem Bioscience Inc., Philadelphia, PA). The Pharmaceutical Development Service, Clinical Center, National Institutes of Health (Bethesda, MD), prepared the injections for use in a blinded manner and also randomly determined the order of treatments for each patient. During both treatment phases, patients took a 2mg dose of oral micronized E 2 daily (Estrace; Mead Johnson Laboratories, Evansville, IN) and a 10-mg/d dose of medroxyprogesterone acetate from the 1st to the 14th calendar day of each month (Provera; The Upjohn Company, Kalamazoo, MI). After each 4-month treatment phase, the injections and hormonal replacement therapy were stopped, and patients were monitored for return of ovarian function. Before they began the second treatment phase, patients returned to the National Institutes of Health Clinical Center, and we reconfirmed hypergonadotropic amenorrhea. 52
Nelson et al.
Premature ovarian failure
Outcome
After each treatment phase, patients had weekly serum samples drawn for 8 weeks; the serum was stored frozen and subsequently sent to us for assay. We measured serum E 2 , progesterone (P), FSH, and luteinizing hormone (LH) by radioimmunoassay methods reported previously (21). We prospectively defined a serum E 2 > 50 pg/mL (184 pmoljL) as representing ovarian follicle function (22) and a serum P > 3 ng/mL (9.5 nmol/L) as representing ovulation (23). For statistical analysis we used Wilcoxon's two-sample test, Fisher's exact test, and X 2 as appropriate; we set P < 0.05 to be significant.
RESULTS Twenty-three women completed both phases of the study. One woman withdrew for personal reasons, one woman withdrew after she developed menometrorrhagia while taking E replacement therapy, and one woman withdrew while taking placebo injections because of the acute onset of newly diagnosed psoriatic arthritis requiring therapy. Deslorelin plus E replacement therapy suppressed serum FSH and LH levels to significantly lower levels (FSH 5.8 ± 1.6, LH 11 ± 0.9) (mean ± SEM) than placebo plus E/progestin replacement therapy (FSH 66 ± 9.6, LH 45 ± 6.7) (P < 0.0001) (Fig. 1). After finishing deslorelin and E replacement ther-
150 140 130 120 110 100
...2 .....
•
8 0
o FSH
eLH 0
©
90 80 70 80 50 40 30 20 10
0
~ 0
• ••
•
• __._
0 0
§ 8 0
0 Placebo
I
•I
0 0
•
+
Drug
Figure 1 Values of FSH and LH at the end of the placebo phase (placebo injections and E replacement therapy) and at the end of the treatment phase (GnRH-a injections combined with E replacement therapy). The horizontal lines represent the respective means for the 23 women who completed the study. The means are significantly different (P < 0.0001).
Fertility and Sterility
apy, five women (22%) demonstrated ovarian follicle function, as compared with eight women (35%) with ovarian follicle function after placebo and E replacement therapy (Fig. 2). After finishing deslorelin and E replacement therapy, three women (13%) ovulated as compared with two women (9%) who did so after placebo and E replacement therapy (Fig. 3). Two patients had ovarian follicle function after both phases, and one of these patients ovulated after both phases. Both patients had clearly demonstrated ovarian failure before beginning the second phase. Deslorelin combined with E replacement therapy did not significantly increase the proportion of patients with ovarian follicle function or ovulation despite the more profound gonadotropin suppression obtained with this therapy. Overall, 11 patients (48%) demonstrated ovarian follicle function and 4 of these patients (17%) ovulated during this study. Patients with 2 or more years since the time of their diagnosis fared no worse than patients with
