Use of Crinone∗ vaginal progesterone gel for luteal support in in vitro fertilization cycles∗∗Crinone 8% (90 mg), Columbia Research Laboratories, Miami, Florida

FERTILITY AND STERILITY威 VOL. 72, NO. 5, NOVEMBER 1999 Copyright ©1999 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A.

IN VITRO FERTILIZATION

Use of Crinone* vaginal progesterone gel for luteal support in in vitro fertilization cycles Samuel J. Chantilis, M.D., Khaled M. Zeitoun, M.D., Snehal I. Patel, B.S., D. Alan Johns, M.D., Valerie A. Madziar, R.N., and Donald D. McIntire, Ph.D. Department of Obstetrics and Gynecology, The University of Texas Southwestern Medical Center, Dallas, Texas

Received November 16, 1998; revised and accepted June 1, 1999. Supported in part by Columbia Research Laboratories, Inc., Rockville Center, New York. Presented at the 54th Annual Meeting of the American Society for Reproductive Medicine, San Francisco, California, October 4 –9, 1998. Reprint requests: Samuel J. Chantilis, M.D., 8160 Walnut Hill Lane, Suite 320, Dallas, Texas 75231 (FAX: 214-373-0217) * Crinone 8% (90 mg), Columbia Research Laboratories, Miami, Florida. 0015-0282/99/$20.00 PII S0015-0282(99)00362-3

Objective: To investigate the efficacy and safety of intravaginal Crinone 8% (Columbia Research Laboratories, Miami, FL) versus IM progesterone for luteal phase support after IVF-ET. Design: Prospective open trial with comparison to historical controls. Setting: University hospital. Patient(s): Two hundred six women undergoing IVF-ET. Intervention(s): One hundred patients received Crinone vaginal progesterone gel (90 mg once daily) and 106 patients received IM progesterone (50 mg once daily) beginning on the evening of oocyte retrieval. Main Outcome Measure(s): Pregnancy and miscarriage rates, and midluteal serum progesterone levels. Result(s): Positive ␤-hCG pregnancy rates, clinical pregnancy rates per transfer, and ongoing pregnancy rates were similar for the Crinone and IM progesterone groups. Women who received Crinone had higher rates of biochemical pregnancy loss but lower rates of clinical pregnancy loss (i.e., spontaneous abortion) than women who received IM progesterone. Midluteal serum progesterone concentrations were significantly higher in the IM progesterone group (94.3 ⫾ 8.8 ng/mL versus 57.7 ⫾ 7.4 ng/mL). Several women who received Crinone had vaginal bleeding 11–13 days after oocyte retrieval. Conclusion(s): For all age categories, positive ␤-hCG and ongoing pregnancy rates were similar when Crinone or IM progesterone was given for luteal phase support in IVF-ET cycles, despite lower serum progesterone concentrations and higher rates of biochemical pregnancy loss with Crinone. Although the results of this study support the use of Crinone as an acceptable alternative for luteal support after IVF-ET, differences in bleeding patterns and rates of biochemical pregnancy loss demonstrate the need for a prospective randomized study. (Fertil Steril威 1999;72:823–9. ©1999 by American Society for Reproductive Medicine.) Key Words: Crinone, progesterone gel, intramuscular progesterone, in vitro fertilization, embryo transfer

Progesterone supplementation during the luteal phase of IVF cycles has become a standard procedure (1). Progesterone, the natural hormone produced by the corpus luteum, is responsible for the activation of endometrial transformations that are indispensable for implantation and maintenance of pregnancy. Luteal phase deficiency after IVF may occur because of persistent inhibition of LH after the use of a GnRH agonist (GnRH-a) for pituitary down-regulation and/or granulosa cell removal after multiple follicle aspiration. Progesterone supplementation during the luteal phase of IVF cycles has been demonstrated to improve pregnancy rates, particularly when a GnRH-a is used (1). Progesterone is an oil-soluble hormone that

can be administered by the IM, vaginal, oral, nasal, rectal, and transdermal routes. Some of these routes are impractical because of obvious inconvenience and/or poor absorption and insufficient bioavailability. Albeit effective, the vaginal route traditionally has been a problem because of the messy vaginal discharge that is associated with the use of cocoa butter– or polyethylene glycol– based progesterone suppositories. Until recently, a vaginal progesterone preparation manufactured by a pharmaceutical company has not been available in the United States. Progesterone suppositories prepared by pharmacists through a practice referred to as compounding may contain variable doses of progesterone because of a lack of quality control. Thus, IM progesterone has be823

come the most widely used option in the United States, despite its significant discomfort. In August 1997, Crinone 8% (90 mg, wt/wt; Columbia Research Laboratories, Miami, FL) became available after approval by the U.S. Food and Drug Administration for use in assisted reproductive technology programs. Crinone 8% is natural progesterone contained in a polycarbophil-based vaginal gel (1.125 g). Two randomized trials have been published regarding the use of Crinone with assisted reproductive techniques. In a multicenter, randomized, prospective trial (2), Crinone 8% was found to be equally effective as oral micronized progesterone for luteal phase support during IVF cycles. In a second study (3), Crinone was evaluated in an oocyte donation model that compared traditional IM progesterone with Crinone 8% (90 mg, wt/wt), administered twice daily in a randomized, prospective fashion. No differences in ongoing pregnancy rates were noted, and patient acceptance of Crinone seemed favorable. In an effort to ease the burden of IM progesterone administration experienced by many of our IVF patients, we conducted a study that compared Crinone with IM progesterone for luteal phase support during an IVF cycle.

MATERIALS AND METHODS Study Demographics Patients scheduled to undergo an IVF cycle beginning in August 1997 were screened for eligibility to enter the study. Patients were excluded from participation if significant genitourinary disease was diagnosed or if they were referred from a clinic in a different city where follow-up treatment was impractical. Eligible patients had a Papanicolaou smear that did not preclude attempted pregnancy within 1 year of beginning the study. Patients excluded from the study were those with a history of dysfunctional uterine bleeding, current urogenital disease, or previous allergic response to a progesterone product. Further, patients were not permitted to use any other progesterone or progestin drug product during the IVF treatment cycle. A patient was allowed to enter the study only once and was excluded during subsequent IVF cycles. The study protocol and consent form were approved by our institutional review board. By design, the study was terminated after 100 consecutively seen patients had been enrolled because our center was only allotted a maximum of 100 study enrollees. The IM progesterone group (controls) consisted of all patients who underwent IVF using fresh, homologous oocytes in 1996. We used the data reported to the Society for Assisted Reproductive Technology (SART), which was modified to report ongoing pregnancy rates as opposed to delivery rates for the purpose of making similar comparisons.

Treatment Protocols Most of the patients in the Crinone and IM progesterone study groups received ovarian stimulation using our standard 824

Chantilis et al.

Crinone for luteal support IVF

luteal phase GnRH-a down-regulation protocol. The administration of leuprolide acetate (0.5–1 mg daily; TAP Pharmaceuticals Inc., Deerfield, IL) was initiated in the luteal phase and continued until ovarian suppression was confirmed with E2 levels and ultrasound examination. The dosage then was reduced by half when ovarian stimulation was commenced. Ovarian stimulation consisted of an FSH-weighted mixture of FSH and hMG, typically at an initial starting dosage 3– 6 ampules per day. Stimulation was altered in accordance with the findings on follicular monitoring and continued until at least two follicles with a mean diameter of ⱖ18 mm were demonstrated on ultrasound examination. At that time, 10,000 U of hCG (Serono Laboratories, Norwell, MA) was administered to all patients with peak E2 levels of ⬍4,000 pg/mL. Five thousand units was administered to patients with peak E2 levels of ⬎4,000 pg/mL on the day of hCG administration. Oocyte retrieval was performed 35–36 hours after hCG administration. Patients designated as poor responders received a “microflare” regimen. This was a different stimulation regimen that consisted of leuprolide acetate (20 ␮g given subcutaneously twice daily starting on cycle day 3) followed by 3– 6 weeks of an oral contraceptive. On cycle day 5, the administration of gonadotropins was initiated at a dosage of 6 ampules per day while the microdose leuprolide acetate regimen was continued. The remainder of the microflare protocol was similar to our routine protocol. Stimulation protocols at the University of Texas Southwestern IVF Clinic remained largely unchanged from 1996 through August 1998. Most of the Crinone and control cycles were luteal phase leuprolide acetate down-regulation protocols conducted by one physician (S.J.C.). Hence, most of the patients in the Crinone and control groups received similar treatments. However, a few changes affected patients in the Crinone group. The new microflare protocol that was prescribed for designated poor responders or patients who had one prior canceled cycle was introduced in our clinic in August 1996. Therefore, patients treated before August 1996 were not candidates for the protocol. Also, one patient in the Crinone group underwent a protocol that included clomiphene citrate and FSH. Further, day 5 ETs introduced in April 1998 for patients with ⱖ6 grade A or B embryos on day 2 affected only the Crinone group. Last, our intracytoplasmic sperm injection (ICSI) program was initiated in 1996, and the proportion of couples who underwent ICSI was compared.

Analysis The outcome of IVF-ET in patients who received luteal support with vaginal progesterone gel (Crinone group) was compared with that of patients who received IM progesterone (historical controls). For the purposes of our study, a positive ␤-hCG level was defined as ␤-hCG level of ⬎5 Vol. 72, No. 5, November 1999

TABLE 1 Demographic characteristics of the patients in the Crinone and IM progesterone groups. Patient group

Parameter Mean (⫾SEM) patient age (y) Mean (⫾SEM) no. of embryos transferred* No. (%) of patients with indicated cause of infertility* Endometriosis Male factor Ovulatory factor Tubal factor Idiopathic No. (%) of patients who underwent indicated stimulation protocol* Standard Microflare Clomiphene citrate No. (%) of patients with indicated day of ET* Day 2–3 Day 5 No. (%) of patients who underwent ICSI*

Crinone (n ⫽ 100)

IM progesterone (n ⫽ 106)

34.2 ⫾ 2.1 3.7 ⫾ 1.1

33.5 ⫾ 2.2 3.5 ⫾ 1.1

19 (19) 20 (20) 25 (25) 27 (27) 9 (9)

16 (12.7) 33 (26.2) 28 (22.2) 43 (34.1) 6 (4.8)

87 (87) 12 (12) 1 (1)

101 (95) 5 (5) 0

94 (94) 6 (6) 30 (30)

106 (100) 0 22 (20.8)

Note: ICSI ⫽ intracytoplasmic sperm injection. * Pⱖ0.5. Chantilis. Crinone in IVF cycles. Fertil Steril 1999.

mIU/mL, and a biochemical pregnancy loss was defined as a positive ␤-hCG level in a patient who failed to achieve a clinical pregnancy. A clinical pregnancy was defined by the presence of a gestational sac, with or without fetal heart movements, on ultrasound examination. Spontaneous abortion was defined as the loss of a clinical pregnancy before 12 weeks’ gestation. An ongoing pregnancy was defined by the identification of fetal heart movements at ⬎12 weeks’ gestation. Midluteal progesterone levels were obtained 6 –7 days after oocyte retrieval in the control and study groups. Postretrieval midluteal progesterone levels obtained during the first 6 months for the IM progesterone group were used for comparison with the study group. Progesterone levels were measured on the day serum was obtained using a chemiluminescent immunoassay (Immulite; Diagnostic Products Corporation, Los Angeles, CA) with a sensitivity of 0.2 ng/mL and a specificity for progesterone of 100%. The intra-assay and interassay coefficients of variation were 5.8%– 8.1% and 7.2%–10%, respectively, for progesterone concentrations of ⬎1 ng/mL. Results obtained in the two treatment groups were analyzed using a Student’s t-test or, in the case of categorical data, the Pearson ␹2 test or Fisher’s exact test when appropriate. All P values are two-sided, and P⬍.05 was considered statistically significant.

RESULTS The demographic characteristics of the patients in the Crinone and IM progesterone groups are presented in Table FERTILITY & STERILITY威

1. There were no differences in mean age, number of embryos transferred, cause of infertility, stimulation protocol, or day of ET between the Crinone and IM progesterone groups. Further, there were no differences in mean age or number of embryos transferred in any of the age categories studied (i.e., ⬍35, 35–39, and ⬎39 years of age) (data not shown). A higher proportion of women who received Crinone underwent ICSI compared with women who received IM progesterone (30% versus 20.8%, P⫽.127), but the difference was not statistically significant. Twenty-two (20.8%) ICSI procedures that resulted in ET were conducted in the control group, compared with 30 (30%) in the Crinone group. Three patients in the control group had a combination of both IVF and ICSI but were included as ICSI patients because at least one embryo transferred had been created using ICSI. Although 33 women who received Crinone ultimately underwent ICSI, 3 cases were conducted on day 2 and were excluded on that basis. The clinical pregnancy rates resulting from ICSI were similar in both groups: 40% (12/ 30) for the Crinone group and 45% (10/22) for the IM progesterone group. From August 1997 through August 1998, a total of 100 patients consented to use Crinone for luteal phase support during IVF-ET. This total included 3 patients (none of whom became pregnant) who were switched to IM progesterone during the midluteal phase because of the personal preferences of the referring physicians. These 3 patients were retained in the Crinone group. Finally, a patient who discon825

TABLE 2 Pregnancy outcome for women who received Crinone vaginal progesterone gel or IM progesterone according to age. Study group Age ⬍35 y Parameter No. of patients with positive ␤-hCG pregnancy/ total no. of patients (%)‡ No. of patients with biochemical pregnancy loss/total no. of patients (%) No. of patients with clinical pregnancy/total no. of patients (%)‡ No. of patients with spontaneous abortion/total no. of patients (%)‡ No. of patients with ongoing pregnancy/total no. of patients (%)‡

Age ⬎39 y

Age 35–39 y

Crinone*

IM progesterone

Crinone†

IM progesterone

Crinone

IM progesterone

25/47 (53.2)

31/63 (49.2)

22/45 (48.9)

16/36 (44.4)

1/8 (12.5)

2/7 (28.6)

0/8 (0)‡

1/2 (50)‡

16/36 (44.4)

1/8 (12.5)

1/7 (14.3)

3/16 (18.8)

1/1 (100)

0/1 (0)

0/8 (0)

1/7 (14.3)

6/25 (24.0)‡

3/31 (9.7)‡

19/47 (40.4)

28/63 (44.4)

2/19 (10.5)

5/28 (17.9)

17/47 (36.2)

23/63 (36.5)

7/22 (31.8)§ 15/45 (33.3) 0/15 (0) 15/45 (33.3)

0/16 (0)§

13/36 (36.1)

* Includes one patient with a cancelled cycle. † Includes two patients with a cancelled cycle. ‡ Pⱖ.05. § P⬍.05. Chantilis. Crinone in IVF cycles. Fertil Steril 1999.

tinued the use of Crinone on day 28 had a biochemical pregnancy diagnosed 2 weeks later. This patient also was retained in the Crinone group. The IM progesterone group consisted of 106 patients who received fresh, homologous oocytes who underwent IVF-ET and used IM progesterone for luteal support. Mean (⫾SEM) midluteal serum progesterone concentrations were significantly higher (P⫽.0015) in the IM progesterone group (94.3 ⫾ 8.8 ng/mL) than in the Crinone group (57.7 ⫾ 7.4 ng/mL). Although bleeding patterns were not a measured outcome, several nonpregnant patients had vaginal bleeding 11–13 days after oocyte retrieval (i.e., 1–3 days before the scheduled day of ␤-hCG testing). The outcome of IVF in both study groups was evaluated for positive ␤-hCG pregnancy rates, biochemical pregnancy losses, clinical pregnancy rates, spontaneous abortion rates, and ongoing pregnancy rates in three age categories: ⬍35, 35–39, and ⬎39 years of age (Table 2). Among women who were ⬍35 years of age at the time of cycle initiation, positive ␤-hCG (53.2% versus 49.2%), clinical pregnancy (40.4% versus 44.4%), and ongoing pregnancy (36.2% versus 36.5%) rates were similar in the Crinone and IM progesterone groups, respectively (Table 2). Among women 35–39 years of age, positive ␤-hCG (48.9% versus 44.4%), clinical pregnancy (33.3% versus 44.4%, P⫽.31), and ongoing pregnancy (33.3% versus 36.1%) rates also were similar in the Crinone and IM progesterone groups, respectively (Table 2). The two groups also were compared with respect to biochemical pregnancy losses and spontaneous abortions (clinical losses). Women who received Crinone had higher rates of biochemical pregnancy loss (24% versus 9.7%, 826

Chantilis et al.

Crinone for luteal support IVF

P⫽.28 in patients ⬍35 years of age; 31.8% versus 0, P⫽.013, in patients 35–39 years of age) than women who received IM progesterone (Table 2). Conversely, once clinical pregnancy was confirmed on ultrasound examination, spontaneous abortion occurred less frequently in the Crinone group (Table 2). Hence, although the ongoing pregnancy rates were similar, pregnancy losses appeared to occur earlier in the Crinone group. Among women ⱖ39 years of age, no meaningful differences were seen in pregnancy outcome parameters (Table 2). Because pregnancy losses appeared to occur earlier in the Crinone group than in the IM progesterone group, particularly in women 35–39 years of age, implantation rates were evaluated in women ⬍35 years of age and in women 35–39 years of age (Table 3). Because clinical pregnancy was defined as the presence of a gestational sac, regardless of the presence of fetal heart movements on ultrasound examination, a higher number of sacs without fetal heart movements may explain in part the differences in biochemical versus clinical pregnancy losses between the two groups. Indeed, fetal heart movements were documented every time a gestational sac was present in women ⬍40 years of age who received Crinone, whereas eight gestational sacs without fetal heart movements were seen in the IM progesterone group. Implantation rates among women ⬍35 years of age and 35–39 years of age were similar in the Crinone and IM progesterone groups.

DISCUSSION Progesterone supplementation during the luteal phase of IVF cycles has become standard treatment in most centers. Vol. 72, No. 5, November 1999

TABLE 3 Implantation rates for women who received Crinone or IM progesterone according to age. Study group Crinone Parameter No. of gestational sacs No. of gestational sacs with fetal heart movements No. of embryos transferred Implantation rates No. of gestational sacs per no. of embryos transferred (%) No. of gestational sacs per no. of fetal heart movements (%)

Age ⬍35 y 36 36 186 19.4 19.4

IM progesterone Age 35–39 y 28 28 156 17.9 17.9

Age ⬍35 y 43 38 226 19.0 16.8

Age 35–39 y 28 25 121 23.1 20.7

Note: Pⱖ.05. Chantilis. Crinone in IVF cycles. Fertil Steril 1999.

The results of a meta-analysis (1) of randomized trials of luteal phase support during IVF treatment cycles support the routine use of progesterone for luteal support in IVF-ET cycles in which a GnRH-a is used. Although various modes of administration have been used to provide progesterone, the IM route has gained such widespread acceptance that it is becoming the main method used in many reproductive centers in the United States. The major drawback of this approach is the need for daily, painful injections of an oily solution with the potential risk for inflammatory reactions and sterile abscesses, particularly for pregnant patients who are treated with extended luteal support. Transvaginal administration of progesterone is an attractive treatment option because of the ease of administration. Support for the vaginal route of progesterone administration stems from multiple studies that have demonstrated the efficacy of vaginal administration of micronized progesterone to induce a secretory transformation of the endometrium in patients without ovarian function. In 1990, Bourgain et al. (4) compared the oral, vaginal, and IM administration of progesterone with respect to endometrial histologic findings on day 21 (day 8 of progesterone therapy) in 43 women with primary ovarian failure who underwent 75 hormone-substituted cycles. The vaginal application of 200 mg of micronized progesterone every 8 hours was adequate to induce secretory transformation of the endometrium, whereas the use of oral progesterone (100 mg every 8 hours) and IM progesterone (50 mg twice daily) resulted in an inadequate response and heterogeneous endometria, respectively. Miles et al. (5) compared the pharmacokinetics and endometrial tissue levels of progesterone after administration by the vaginal and IM routes. In their study, 20 agonadal subjects received 200 mg of transvaginal micronized progesterone every 6 hours or 50 mg of IM progesterone-in-oil twice daily. Midluteal (day 7 of progesterone therapy) endometrial biopsy results obtained from women who received vaginal progesterone showed similar histologic dating and FERTILITY & STERILITY威

estrogen and progesterone receptor content as those from women who received IM progesterone. Further, intravaginal progesterone resulted in a higher endometrial progesterone concentration (11.5 ng/mg of protein) compared with that achieved with IM progesterone (1.4 ng/mg of protein). Vaginal progesterone also has been used for oocyte donation treatment cycles that have resulted in respectable rates of pregnancy (6), a condition that obviates secretory transformation of the endometrium. In addition to noncommercial vaginal progesterone preparations, Crinone has been tested in women deprived of ovarian function. Fanchin et al. (7) administered Crinone 8% (containing 90 mg of progesterone) and two other doses of Crinone (4% [45 mg] and 8% [180 mg]) to estrogen-primed women before they underwent a midluteal phase endometrial biopsy. Secretory transformation of the endometrium was achieved with all three doses. Intravaginal micronized progesterone also has been administered during IVF treatment cycles as a form of luteal phase support. Smitz et al. (8) compared 600 mg/d of intravaginal micronized progesterone with 50 mg/d of IM progesterone-in-oil in a prospective, randomized study of 262 women who underwent ovulation induction for IVF after ovarian suppression using a GnRH-a. Their results showed that a higher clinical pregnancy rate was achieved with the use of vaginal progesterone (33.6%) than with the use of IM progesterone (26.7%), but the difference was not statistically significant (P⫽.07). Further, their results showed that the use of vaginal progesterone resulted in lower first trimester abortion rates among women who were treated with the intravaginal micronized progesterone, compared with those who were treated with IM progesterone. A progesterone gel, Crinone 8% (90 mg) was specially designed for vaginal use and became available in the United States in August 1997. Because of its controlled and sustained-release properties, this gel limits the need for appli827

cation to once daily. It is particularly attractive because it is easy to administer and obviates the need for IM injections. Two published studies have supported the use of this vaginal progesterone gel in assisted reproduction. Pouly et al. (2) studied Crinone 8% (90 mg) compared with oral micronized progesterone (300 mg; Laboratories Besins Isovesco, Paris, France). In their study, luteal support was initiated within 24 hours after ET. These investigators found that Crinone was equally effective with respect to pregnancy rates, spontaneous abortion rates, and delivery rates as oral micronized progesterone. Gibbons et al. (3) used an oocyte donation model to compare IM progesterone with Crinone 8% administered twice daily and found similar pregnancy rates with these two treatments. Although Crinone 8% has been used effectively in oocyte donation protocols and shown to be equivalent to oral micronized progesterone, a study comparing it with IM progesterone is needed to evaluate the efficacy of this new vaginal progesterone gel for luteal phase support. Micronized vaginal progesterone (600 mg/d), and IM progesterone (50 mg/d) have been demonstrated (8) to produce equivalent pregnancy outcomes, but the new vaginal progesterone gel (i.e., Crinone 8%, containing 90 mg of progesterone and administered daily) has not yet been shown to be equivalent to IM progesterone. The design of this study was descriptive in nature in an effort to provide a preliminary comparison between Crinone and IM progesterone for luteal support for IVF-ET. Although they were not randomized, the two study groups were similar with respect to number of patients, age of patients, number of embryos transferred, cause of infertility (primary diagnosis), stimulation protocol, and day of ET. A higher proportion of patients underwent ICSI in the Crinone group (30%) than in the IM group (20.8%), but the pregnancy rates associated with ICSI in both groups were equivalent. This difference is attributed to the evolution of ICSI at IVF programs; as technical competence with the ICSI procedure progresses, the criteria for ICSI usually are broadened, which results in a greater proportion of patients undergoing the procedure. The size of the study was limited by the number of patients allotted to our center. Assuming a 35% ongoing pregnancy rate, as observed in the IM progesterone (control) group, 328 subjects would be required in each arm to detect a difference of 10% between the two groups. In the present trial, we found no differences in positive ␤-hCG or ongoing pregnancy rates between women who received Crinone and women who received IM progesterone. Differences were noted in bleeding patterns, specifically the onset and type of pregnancy losses among women who received Crinone compared with those who received IM progesterone. Biochemical pregnancy loss occurred more frequently among women who received Crinone, whereas spontaneous abortion (clinical pregnancy loss) occurred less 828

Chantilis et al.

Crinone for luteal support IVF

frequently, which ultimately resulted in similar ongoing pregnancy rates. Although the number of patients 35–39 years of age in this study was limited, the rate of biochemical pregnancy loss in women who received Crinone (31.8%) was statistically different (P⫽.013) than that in women who received IM progesterone (zero). Women ⬍35 years of age who received Crinone also had a higher rate of biochemical pregnancy loss than those who received IM progesterone (24% versus 9.7%, P⫽.28), but the difference was not statistically significant. Although our outcome criteria did not include the day of onset of vaginal bleeding, some patients reported bleeding that occurred before the day of serum ␤-hCG testing. These patients reported vaginal bleeding occurring within 3 days before ␤-hCG testing, which is performed 14 days after oocyte retrieval in our program. Most of these patients were not pregnant, but at least two patients with “premature” bleeding had biochemical pregnancy losses. Although they were not recorded for study purposes, bleeding patterns among Crinone users were noticeably different (i.e., bleeding occurred earlier compared with that seen with IM progesterone). The pattern of earlier bleeding noted in some patients who received Crinone is consistent with the observed trend toward earlier pregnancy loss (i.e., more biochemical pregnancies and fewer clinical spontaneous abortions). Although implantation rates for both groups were similar, women who received IM progesterone had more gestational sacs without fetal heart movements seen on ultrasound examination, a finding that suggests that IM progesterone prolongs the presence of nonviable gestational sacs (i.e., clinical pregnancies [SART definition] without fetal heart movements). Pasquale et al. (9) recently compared daily doses of 100, 200, and 400 mg of micronized progesterone administered vaginally to estrogen-primed postmenopausal women in a randomized, three-way crossover study. Twenty healthy women underwent histologic dating after endometrial biopsy was performed on day 25. All three doses yielded histologic dating consistent with cycle days 25 through 28, but the 100-mg progesterone dose yielded less consistent secretory conversion and earlier bleeding patterns compared with the 200-mg and 400-mg doses. Although earlier bleeding patterns are disconcerting for both the patient and the physician, the ongoing pregnancy rates ultimately were similar between our Crinone group and a historical control group. A randomized, prospective trial comparing Crinone with IM progesterone in women undergoing routine IVF (using homologous oocytes) has not yet been published. In the present trial, we found that midluteal progesterone levels obtained 6 –7 days after oocyte retrieval were markedly lower in patients who received Crinone 8% than in patients who received IM progesterone (historical controls). This finding is in accordance with expectations drawn from Vol. 72, No. 5, November 1999

prior experience. We expected to find the same difference between the Crinone and IM progesterone groups as reported by Gibbons et al. (3) in women who were deprived of ovarian function. Our results supported this assumption. Ultimately, we concluded that progesterone level determinations cannot be used to assess the efficacy of treatment with Crinone or IM progesterone and thus need not be performed. We found no statistically significant differences in positive ␤-hCG and ongoing pregnancy rates in women who received luteal phase support with Crinone or IM progesterone. Women who received Crinone had earlier bleeding patterns associated with higher rates of biochemical pregnancy loss but lower rates of clinical pregnancy loss. However, this may have no relevance to pregnancy outcome. Similarly, plasma progesterone levels are not predictive of pregnancy outcome and thus are of no use in determining efficacy when progesterone is used for luteal support in IVF-ET cycles. Although the results of this descriptive study encourage the use of Crinone for luteal phase support after IVF-ET, they demonstrate the need for a large, prospective, randomized trial to confirm that the different bleeding patterns have no effect on pregnancy outcome.

FERTILITY & STERILITY威

References 1. Soliman S, Daya S, Collins J, Hughes EG. The role of luteal phase support in infertility treatment: a meta-analysis of randomized trials. Fertil Steril 1994;61:1068 –76. 2. Pouly JL, Bassil S, Frydman R, Hedon B, Nicollet B, Prada Y, et al. Luteal support after in-vitro fertilization: Crinone 8%, a sustained release vaginal progesterone gel, versus Ultrogestan, an oral micronized progesterone. Hum Reprod 1996;11:2085–9. 3. Gibbons WE, Toner JP, Hamacher P, Kolm P. Experience with a novel vaginal progesterone preparation in a donor oocyte program. Fertil Steril 1998;69:96 –101. 4. Bourgain C, Devroey P, Van Waesberghe L, Smitz J, Van Steirteghem AC. Effects of natural progesterone on the morphology of the endometrium in patients with primary ovarian failure. Hum Reprod 1990;5:537– 43. 5. Miles RA, Paulson RJ, Lobo RA, Press MF, Dahmoush L, Sauer MV. Pharmacokinetics and endometrial tissue levels of progesterone after administration by intramuscular and vaginal routes: a comparative study. Fertil Steril 1994;62:485–90. 6. Frydman R, Letur-Ko˘nirsch H, de Ziegler D, Bydlowski M, Raoul-Duval A, Selva J. A protocol for satisfying the ethical issues raised by oocyte donation: the free, anonymous and fertile donors. Fertil Steril 1990;53: 666 –72. 7. Fanchin R, De Ziegler D, Bergeron C, Righini C, Torrisi C, Frydman R. Transvaginal administration of progesterone. Obstet Gynecol 1997;90: 396 – 401. 8. Smitz J, Devroey P, Faguer B, Bourgain C, Camus M, Van Steirteghem AC. A prospective randomized comparison of intramuscular or intravaginal natural progesterone as a luteal phase and early pregnancy supplement. Hum Reprod 1992;7:168 –75. 9. Pasquale SA, Foldesy RG, Levine JP, Bachmann GA, Blackwell RE. Peripheral progesterone (P) levels and endometrial response to various dosages of vaginally administered P in estrogen-primed women. Fertil Steril 1997;68:810 –5.

829