REPRODUCTIVE ENDOCRINOLOGY

Transcription

1 FERTILITY AND STERILITY VOL. 76, NO. 5, NOVEMBER 2001 Copyright 2001 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A. REPRODUCTIVE ENDOCRINOLOGY The effect of nonsteroidal antiinflammatory drugs on ovulation: a prospective, randomized clinical trial Meike L. Uhler, M.D., a Judith W. Hsu, M.D., a Susan G. Fisher, Ph.D., b and Michael J. Zinaman, M.D. a Loyola University Medical Center, Maywood, Illinois Received January 26, 2001; revised and accepted May 18, Presented in part at the Annual Meeting of the American Society for Reproductive Medicine, San Diego, California, October 21 26, Sponsored in part by the Vioxx Medical School Grant Program at Merck & Co., Inc., Whitehouse Station, NJ. Reprint requests: Meike L. Uhler, M.D., Loyola University Medical Center, Department of Obstetrics and Gynecology, 2160 South First Avenue, Maywood, Illinois (FAX: ; mulher@lumc.edu). a Division of Reproductive Endocrinology and Infertility, Department of Obstetrics and Gynecology. b Department of Medicine /01/$20.00 PII S (01) Objective: To assess the effect of ibuprofen, a nonspecific inhibitor of prostaglandin synthesis, on ovulation. Design: Prospective, randomized, double-blind, placebo-controlled cross-over study. Setting: University Medical Center. Patient(s): Twelve normally cycling women between ages 20 and 40. Intervention(s): Subjects were randomized to either oral ibuprofen (800 mg) or placebo three times per day, beginning when the maximum diameter of the leading follicle reached 16 mm by ultrasound, and continuing for 10 days total. The second cycle was a washout period, and in the third cycle, the subjects were crossed over to the alternate regimen from the first cycle. The probability of delayed follicular collapse was determined using the binomial distribution, and changes in P levels were compared using the paired t test. Main Outcome Measure(s): Urinary LH surge, follicular collapse by serial transvaginal ultrasonography, and serum midluteal P levels. Result(s): Eleven of 12 subjects detected an LH surge with both ibuprofen and placebo. Five of 11 women demonstrated a 2-day increase in time interval from detection of the LH surge to follicular collapse, and 3 of those 5 had been randomized to ibuprofen. This represents a 27% (3 of 11; 95% confidence limits: 1%, 53%) rate of delay for follicular collapse for ibuprofen. There was no difference in average midluteal P levels for ibuprofen or placebo. Conclusion(s): If ibuprofen inhibits follicular collapse, this effect is seen in a small group of study subjects, and this information should be clinically reassuring to patients who take nonsteroidal anti-inflammatory drugs. Serum midluteal P levels were unaffected by administration of ibuprofen. (Fertil Steril 2001;76: by American Society for Reproductive Medicine.) Key Words: Follicular rupture, ultrasonography, ovary, ibuprofen, urinary LH test, progesterone There is compelling evidence that prostaglandins play an important role in ovulation, specifically in the process of follicular rupture and collapse (1 4). In animals, the administration of nonsteroidal anti-inflammatory drugs has been shown to block the synthesis of prostaglandins (5, 6), inhibit ovulation (7, 8), prevent follicular collapse (9), and produce other effects that are similar to those seen in the luteinized unruptured follicle syndrome (10). Although many studies have been done in animal models (11, 12), there are few data on the inhibition of human ovulatory function by this class of drugs (13, 14). Many reproductive-age women routinely take this medication for a variety of ailments, including dysmenorrhea and pelvic discomfort. In addition, infertility patients undergoing ovulation induction, including many suffering from endometriosis, are often advised to avoid such medications because ovulation may be inhibited. This investigation was undertaken to determine the effect of the administration of a popular medication, ibuprofen, on ovulation in normally cycling women. This study has broad implications for clinical patient management and the process of natural ovulation. MATERIALS AND METHODS Subjects Subjects were recruited by local advertisement. Inclusion criteria included women between the ages of 20 and 40 years within 20% 957

2 of their ideal body weight with regular menstrual cycles. The subjects were required to be sexually abstinent, to be contracepting by a barrier method, or to have had permanent sterilization. Subjects had not been using any type of hormonal therapy with the exception of thyroid replacement or been using any nonsteroidal anti-inflammatory agents for 2 months before study entry. Exclusion criteria included current or past peptic ulcer disease, gastritis, or sensitivity to nonsteroidal anti-inflammatory drugs; abnormal vaginal bleeding; cervical dysplasia; or significant medical illnesses including diabetes mellitus, hypertension, coronary artery disease, cancer, thromboembolic disease, liver disease, renal disease, or lung disease. All participants underwent a complete history and physical examination with a Papanicolaou smear, unless one had been performed within the previous 12 months. Subjects were required to be ovulatory by meeting two of the three following criteria: detection of a positive urinary LH surge, serial transvaginal ultrasonography to document follicular growth and collapse, and serum midluteal P level of 5 ng/ml during the month leading into intervention. Design Using a double-blind, cross-over study design, all subjects were randomized by computer-generated random number to receive either ibuprofen or placebo in the first treatment cycle. Ibuprofen was given orally in the form of a liquid suspension (800 mg [40 ml] three times daily), to begin when the mean maximum diameter of the leading follicle reached 16 mm according to transvaginal ultrasonography. The timing of the first ultrasound was calculated by subtracting 18 from the shortest length of the previous three menstrual cycles. The placebo medication was composed of a red-colored diluent in a suspending agent with yellow food coloring made up to match the study medication, and 40 ml was taken three times daily. Each was administered for 10 days. At the time that ibuprofen or placebo was started, the subjects also used a daily urinary LH kit each morning until an LH surge was detected and underwent serial transvaginal ultrasonography in the morning until follicular collapse was documented. The second cycle was a washout menstrual cycle. In the third cycle of study, the subjects were crossed over to the alternate regimen from the first cycle. The primary endpoint for each subject was defined as a 2-day increase in time from detection of the LH surge to follicular collapse with ibuprofen-treated cycles as compared with placebo cycles. The study was approved by the institutional review board at Loyola University Medical Center, and each woman gave written informed consent. Serial transvaginal ultrasonography was performed (General Electric, Logiq 400 MD, Milwaukee, WI) until follicular collapse was documented. Seven and nine days after the detection of urinary LH surge (with Clearplan, Unipath Ltd., Bedford, United Kingdom), blood samples were drawn for P. The samples of peripheral blood were centrifuged, and the serum was removed for hormonal analysis and stored at 70 C. Serum P levels were measured by chemiluminescence assay (ACS:180, Chiron Diagnostics Corp., East Walpole, MA), with an interassay CV of 10%. Samples from all subjects were analyzed in two batches to minimize intraassay variability. Statistical Analysis The sample size estimate was based on the assumption that 40% of the subjects would experience a 2-day increase in time from detection of the LH surge to follicular collapse during an ibuprofen-treated cycle compared with a placebotreated cycle. It was assumed that 10% of the subjects would have such a delay with placebo cycles (15, 16) compared with ibuprofen cycles. Using a one-sample test of proportions (40% compared with population estimate of 10%), it was determined that 11 subjects would be required to achieve a power of 80% with a two-sided alpha level of For all statistical testing, a two-sided alpha level of 0.05 was considered statistically significant. The probability of delayed follicular collapse between ibuprofen and placebo was determined using the binomial distribution. The average midluteal serum P levels between ibuprofen and placebo cycles were compared using the paired t test. RESULTS A total of 24 women were recruited into the trial, 23 women were screened, and 15 women passed the screening phase and were randomized. Three of the 15 randomized women did not complete the study for the following reasons: one moved away, one became pregnant during the washout month, and one was traveling during a study month. A total of 12 subjects completed the study. Eleven of 12 subjects detected a positive LH surge with both ibuprofen and placebo. One subject did not detect an LH surge during month 3 of the study and therefore was not included in the data analysis. The total number of evaluable subjects was 11, and the mean SD of their ages was 31 6 years. In Table 1, each subject s data are presented for cycle day of LH surge and follicular collapse and the number of cycle days from LH surge to follicular collapse. Five of 11 women demonstrated a 2-day increase in time interval from detection of the LH surge to follicular collapse in cycle 1 compared with cycle 3 or vice versa. Three of the five subjects who showed a delay in follicular collapse were randomized to ibuprofen. This represented a 27% (3 of 11; 95% confidence limits [CL]: 1%, 53%) rate of delay for follicular collapse on ibuprofen. The background rate of delay for placebo was 18% (2 of 11) in this study. The mean midluteal serum P levels (averaged between cycle day 7 and 9 for each subject) for ibuprofen and placebo cycles were (95% CL: 9.6, 12.7) and (95% CL: 8.9, 15.2), respectively. This difference was not statistically significant (paired t 0.71, P.49). 958 Uhler et al. Effect of ibuprofen on ovulation Vol. 76, No. 5, November 2001

3 TABLE 1 Cycle day of LH surge and follicular collapse and number of cycle days from LH surge to follicular collapse for each subject. Subject No. Drug Cycle LH surge (cycle day) Follicular collapse (cycle day) No. of cycle days from LH surge to follicular collapse 1 a Ibuprofen Placebo Ibuprofen Placebo Ibuprofen Placebo Ibuprofen Placebo Ibuprofen Placebo a Ibuprofen Placebo a Ibuprofen Placebo Ibuprofen Placebo Ibuprofen Placebo Ibuprofen Placebo Ibuprofen Placebo a Denotes increase in time interval of 2 days from LH surge to follicular collapse with ibuprofen cycle compared with placebo cycle. Uhler. Effect of ibuprofen on ovulation. Fertil Steril Several subjects reported a dislike of the taste of the liquid study medication; however, no one reported any adverse gastrointestinal side effects. One subject noted ecchymosis on her arm at the phlebotomy site that resolved in 2 weeks. DISCUSSION This is the first randomized, placebo-controlled trial undertaken to determine the effects of the administration of ibuprofen, a nonspecific prostaglandin inhibitor, on the ovulatory process in normally cycling women. Ibuprofen was selected because it is available in an inexpensive generic form over the counter and is readily accessible to most patients. Furthermore, the maximum daily dose of 2,400 mg of ibuprofen was chosen to optimally demonstrate a delay in ovulation if such an effect did indeed exist. Our results suggest that if ibuprofen delayed follicular collapse, this effect was seen in a small group of women, and this information should be clinically reassuring to patients who often take nonsteroidal anti-inflammatory drugs. Our study results confirm those of other investigators who have found that serum midluteal P levels were unaffected by administration of nonsteroidal anti-inflammatory drugs (13, 14). Although our study had 80% power to detect a 40% delay in follicular collapse with ibuprofen, it was still limited by the small sample size. The initial power analysis assumed a 10% follicular delay in the placebo group. This number was chosen because it is thought that nonsteroidal anti-inflammatory drugs prevented follicular collapse, an effect similar to that seen with luteinized unruptured follicle syndrome, which has been estimated at approximately 5% 10% of the population (15, 16). In our trial, we found an 18% follicular delay in the placebo group. This finding would require a larger sample to still detect a 40% delay in the ibuprofen group. Smaller differences in follicular collapse may be clinically important to some women, particularly infertility patients timing intercourse or intrauterine insemination; however, differences in follicular collapse, such as 10% or even 20%, between the treatment and placebo groups would require a considerable increase in sample size. Several factors that could potentially affect ovulation were controlled in this study. All study participants were within 20% of their ideal body weight. Subjects were not allowed to take other medication while participating in the trial, with the exception of thyroid replacement, and all subjects used the same brand of urinary LH kits for detection of the LH surge. Additional factors known to affect ovulation were not controlled in this study, such as initiation of smoking or the effect of stress through major life changes FERTILITY & STERILITY 959

4 such as divorce or death in the family, but the duration of this study was very short, and we are unaware of any such events. A further limitation to the study is that although subjects maintained a daily medication log, serum ibuprofen levels were not obtained; thus, compliance cannot be fully assured. If follicular rupture is truly a process controlled by local factors, a much higher dose of ibuprofen at the site of the ovarian follicle may be required to consistently impede ovulation. There are few human studies available that examine the potential effects of nonsteroidal anti-inflammatory medications on ovulation. Five fertile women were given a 75-mg daily oral dose of indomethacin, and all five women demonstrated ovulation based on a midluteal P level of 4 ng/ml (17). A larger study that included 10 healthy fertile women given oral naproxen (250 mg twice daily) throughout the menstrual cycle found that naproxen did not suppress ovulation, based on a rise in serum LH levels and secretory endometrial biopsies (18). Serial ultrasonic scans were performed in 20 healthy females with regular menstrual cycles given hcg to induce follicle rupture when the mean diameter of the dominant follicle reached 18 mm (19). Luteinized unruptured follicles were seen in 10.7% of untreated cycles. When an oral prostaglandin synthethase inhibitor (azapropazone, 1.2 g twice daily on the first day and then 900 mg twice daily for 4 days, or indomethacin, 100 mg twice daily for 5 days) was administered, the incidence of luteinized unruptured follicles was greatly increased: to 50% with azapropazone and 100% with indomethacin. A larger study of 41 women awaiting tubal sterilization was performed in a randomized, placebo-controlled fashion with indomethacin, bromfenac, or azapropazone administered in the periovulatory period (13). Prostaglandin E 2 and PGF 2 levels in the follicular fluid were significantly reduced by indomethacin and bromfenac compared with placebo, and azapropazone had no effect on follicular fluid eicosanoid concentrations, probably because it is a comparatively weaker nonsteroidal anti-inflammatory drug. There was no difference among the four groups regarding midluteal serum estradiol and P concentrations. A case series of three young women with inflammatory arthritis on nonsteroidal anti-inflammatory drugs showed development of luteinized unruptured follicles (20). More recently, six healthy volunteers awaiting treatment for infertility with in vitro fertilization were given oral indomethacin (50 mg, three times a day) when the leading follicle reached 16 mm or at the time of a positive LH surge (14). Follicular rupture was delayed in five of six patients by 2 to 12 days, and there was no significant effect of indomethacin on the levels of circulating FSH, estradiol, LH, or P or menstrual cycle length. Earlier animal studies implicated the involvement of prostaglandins in the regulation of ovarian follicular function. It was demonstrated that inhibitors of prostaglandin synthesis, such as aspirin and indomethacin, were capable of blocking ovulation in rats (1). These initial findings were later confirmed in various other species including mice, rabbits, rhesus and marmoset monkeys, pigs, and goldfish (1). In rabbits and goldfish, the inhibitor was effective when it was applied locally to the follicle, implying a direct effect at the follicular level rather than a systemically mediated effect such as an inhibition at the level of the hypothalamic-pituitary axis. Further support was added when investigators reported the marked increase in intrafollicular levels of prostaglandins in the E and F series in several animal species shortly before ovulation (3). Although these prostaglandin inhibitors appeared to block ovulation in animal models in a similar fashion as seen in the luteinized unruptured follicle syndrome (10), blockade of ovulation by indomethacin was not accompanied by blockade of P secretion and corpus luteum formation, and it did not appear to interfere with the normal preovulatory LH surge (1, 9). On a cellular level, the suggestion that ibuprofen may delay ovulation in any normally cycling woman lends credence to the notion that ovulation may be controlled by factors arising from the arachidonic acid pathway and that it may be an inflammatory-like process that leads to the targeted rupture of a human follicle (2). The exact effect of ibuprofen on human preovulatory follicular fluid prostaglandins has not been studied, but ibuprofen may reduce follicular fluid levels of PGE 2 and PGF 2, similar to the case with other nonsteroidal anti-inflammatory drugs, namely indomethacin and bromfenac (13). The evidence that circulating levels of hormone such as P, LH, estradiol, and FSH are unchanged in the presence of nonsteroidal anti-inflammatory drugs, whereas ovulation is altered, suggests that local factors may be the predominant driving force to the mechanism of ovulation (14, 19). Because many reproductive-age women take nonsteroidal anti-inflammatory drugs on a regular basis for control of their dysmenorrhea and nonspecific pelvic pain, the results of this investigation may have important implications in clinical management. Our study results suggest that if ibuprofen delays follicular collapse, this is seen in a small group of subjects, and this finding should be reassuring to our patients who often take nonsteroidal anti-inflammatory drugs. Ibuprofen does not affect midluteal P levels in normally cycling women. Smaller effects in delay of follicular collapse caused by nonsteroidal anti-inflammatory drugs may be clinically significant for some women; however, precise estimates of the magnitude of such an effect would require the conduct of a much larger prospective study. Acknowledgments: The authors gratefully acknowledge the gift of Clearplan LH predictor kits from Janet Jacobs, Unipath Ltd., Bedford, United Kingdom. Both the study medication and placebo were supplied by the Investi- 960 Uhler et al. Effect of ibuprofen on ovulation Vol. 76, No. 5, November 2001

5 gational Drug Section of the Pharmacy Department of Loyola University Medical Center. References 1. Armstrong DT. Prostaglandins and follicular functions. J Reprod Fertil 1981;62: Priddy AR, Killick SR. Eicosanoids and ovulation. Prostaglandins Leukot Essent Fatty Acids 1993;49: Lindner HR, Zor U, Kohen F, Bauminger S, Amsterdam A, Lahav M, et al. Significance of prostaglandins in the regulation of cyclic events in the ovary and uterus. Adv Prostaglandin Thromboxane Res 1980;8: Seibel MM, Swartz SL, Smith D, Levesque L, Taymor ML. In vivo prostaglandin concentrations in human preovulatory follicles. Fertil Steril 1984;42; Espey LL, Norris C, Saphire D. Effect of time and dose of indomethacin on follicular prostaglandins and ovulation in the rabbit. Endocrinol 1986;119: Zanagnolo V, Dharmarajan AM, Endo K, Wallach EE. Effects of acetylsalicylic acid (aspirin) and naproxen sodium (naproxen) on ovulation, prostaglandin, and progesterone production in the rabbit. Fertil Steril 1996;65: Katz E, Dharmarajan AM, Sueoka K, Ghodgaonkar RB, Dubin NH, Wallach EE. Effects of systemic administration of indomethacin on ovulation, luteinization, and steroidogenesis in the rabbit ovary. Am J Obstet Gynecol 1989;161: Carlson JC, Barcikowski B, Cargill V, McCracken JA. The blockade of LH release by indomethacin. J Clin Endocrinol Metab 1974;39: Maia H, Barbosa I, Coutinho EM. Inhibition of ovulation in marmoset monkeys by indomethacin. Fertil Steril 1978;29: Murdoch WJ, Cavender JL. Effect of indomethacin on the vascular architecture of preovulatory ovine follicles: possible implication in the luteinized unruptured follicle syndrome. Fertil Steril 1989;51: Miyazaki T, Dharmarajan AM, Atlas SJ, Katz E, Wallach EE. Do prostaglandins lead to ovulation in the rabbit by stimulating proteolytic enzyme activity? Fertil Steril 1991;55: Kranzfelder D, Reich R, Abisogun O, Tsafriri A. Preovulatory changes in the perifollicular capillary network in the rat: role of eicosanoids. Biol Reprod 1992;46: Priddy AR, Killick ST, Elstein M, Morris J, Sullivan M, Patel L, et al. The effect of prostaglandin synthetase inhibitors on human preovulatory follicular fluid prostaglandin, thromboxane, and leukotriene concentrations. J Clin Endocrinol Metab 1990;71: Athanasiou S, Bourne TH, Khalid A, Okokon EV, Crayford T, Hagstrom HG. Effects of indomethacin on follicular structure, vascularity, and function of the periovulatory period in women. Fertil Steril 1996; 65: Kerin JF, Kirby C, Morris D, McEvoy M, Ward B, Cox LW. Incidence of luteinized unruptured follicle phenomenon in cycling women. Fertil Steril 1983;40: Hamilton CJCM, Wetzels LCG, Evers JLH, Hoogland HJ, Muijtjens A, de Haan J. Follicle growth curves and hormonal patterns in patients with the luteinized unruptured follicle syndrome. Fertil Steril 1985;43: Toppozada M, El-Abd M, El-Sokkary H, El-Rahman HA. Effect of a prostaglandin inhibitor of human ovulation. Singapore J Obstet Gynecol 1979;10: Lundstrom V, Landgren BM, Eneroth P, Johannisson E. The effect of a prostaglandin synthetase inhibitor on the hormonal profile and the endometrium in women. Acta Obstet Gynecol Scand Suppl 1983;113: Killick S, Elstein M. Pharmacologic production of luteinized unruptured follicles by prostaglandin synthetase inhibitors. Fertil Steril 1987; 47: Smith G, Roberts R, Hall C, Nuki G. Reversible ovulatory failure associated with the development of luteinized unruptured follicles in women with inflammatory arthritis taking non-steroidal anti-inflammatory drugs. Br J Rheum 1996;35: FERTILITY & STERILITY 961