0021-972X/90/7101-0235$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1990 by The Endocrine Society
Vol. 71, No. 1 Printed in U.S.A.
The Effect of Prostaglandin Synthetase Inhibitors on Human Preovulatory Follicular Fluid Prostaglandin, Thromboxane, and Leukotriene Concentrations* ALVAN R. PRIDDY, STEPHEN R. KILLICK, MAX ELSTEIN, JULIE MORRIS, MARK SULLIVAN, LATA PATEL, AND MURDO ELDER Departments of Obstetrics and Gynaecology (A.R.P., S.R.K., M.E.) and Statistics (J.M.), University Hospital of South Manchester, Manchester, and the Department of Obstetrics and Gynaecology, Hammersmith Hospital (M.S., L.P., M.E.), London, United Kingdom
ABSTRACT. This study evaluates the eicosanoid concentration in luteinized unruptured follicles (LUFs) on the ovaries of patients who had been treated with inhibitors of prostaglandin synthetase. Indomethacin, bromfenac, or azapropazone (or a placebo) was administered orally to 41 women during the periovulatory period. Follicular development was monitored by serial ultrasound examinations, and the onset of ovulation was regulated by an injection of hCG. Follicular fluid was aspirated during sterilization by minilaparotomy, which was performed
just before the expected time of ovulation. Prostaglandin E2 and PGF2a levels in the fluid were significantly reduced by indomethacin and bromfenac compared to those after placebo treatment. Bromfenac also reduced the follicular fluid leukotriene B4 level. Therefore, the development of luteinized unruptured follicles after treatment with nonsteroidal antiinflammatory drugs appears to be associated with a significant decrease in the synthesis of ovarian eicosanoids. (J Clin Endocrinol Metab 7 1 : 235-242, 1990)
P
indomethacin is not overcome by administering LH at doses that would normally induce ovulation (10). Human follicles have been found to contain PGs, thromboxanes (TXs), and leukotrienes (LTs), although most of the data is from the analysis of follicular fluid from assisted reproduction programs. Human follicular fluid from spontaneous cycles has been found to contain PGE2, PGF2a, 6-keto-PGFla, and TXB 2 (11,12), and the follicular fluid PGF2a concentration increases in the immediate preovulatory period (13). We have previously demonstrated that indomethacin can inhibit ovulation {i.e. induce LUF) in women if given at high doses in the periovulatory period (14). We performed serial ultrasonic scans of follicular development throughout 46 spontaneous cycles in 20 healthy female volunteers, and hCG was given to induce follicular rupture on a particular day. LUFs were seen in 10.7% of untreated cycles, but when NSAIDs were administered over the periovulatory period the incidence of LUF was greatly increased (to 100% with indomethacin; n = 8). The present study was carried out to assess whether NSAIDs (PG synthetase inhibitors) reduce follicular fluid PG, TX, and LT concentrations and to assess the relative effects a new NSAID, bromfenac.
ROSTAGLANDINS (PGs) are involved in mammalian ovulation. The cyclooxygenase inhibitor indomethacin inhibits ovulation in nonprimates such as the rat (1), rabbit (2), pig (3), and sheep (4), and in primates such as the monkey (5). This action is due to PG inhibition, as similar results are obtained using antiserum to either PGF2« (6) or to the cyclooxygenase enzyme itself (7). Administration of PGs overcomes this inhibition. PG concentrations in follicular fluid have been shown to increase in the preovulatory period (8). The LH surge initiates protein synthesis in granulosa cells and cyclooxygenase production via cAMP, so there is a delay from this surge to the increase in the PG content of follicular fluid. Administration of indomethacin at the time of the LH surge of hCG administration reduces the concentrations of follicular fluid PGs and inhibits follicle wall rupture. However, other aspects of the ovulatory response, such as progesterone production, are unaffected, resulting in a luteinized, yet unruptured, follicle (LUF). The action is a local one, as an intrafollicular injection of nonsteroidal antiinflammatory drug (NSAID) prevents ovulation (9). Also, inhibition of ovum release by Received August 4, 1989. Address requests for reprints to: Alvan Priddy, M.R.C.O.G., Department of Obstetrics and Gynaecology, University Hospital of South Manchester, West Didsbury, Manchester, M20 8LR, United Kingdom. * This work was supported by A.H. Robbins Ltd.
Subjects and Methods After obtaining ethical committee approval, 41 caucasion female volunteers were recruited from our waiting list for 235
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 November 2015. at 18:17 For personal use only. No other uses without permission. . All rights reserved.
PRIDDY ET AL.
236
JCE & M • 1990 Vol 71 • No 1
50 i
40 -
*I 51p^
30XAA.
1
m m m
1
FIG. 1. Comparison of mean follicular fluid PGE2 concentration and 95% confidence intervals in placebo, azapropazone, indomethacin, and bromfenac groups.
I
Q.
20"
10-
oPLACEBO
AZAPROPAZONE
INDOMETHACIN
BROMFENAC
*group difference p=0.011 sterilization. Informed written consent was obtained from all volunteers. They were 24-40 yr old, had no gynecological abnormalities, and were not using the oral contraceptive pill. They were randomly assigned to 4 groups to be given a NSAID (indomethacin, azapropazone, or bromfenac) or an indistinguishable placebo orally during the periovulatory period. Randomization was achieved by a computor randomization code. Regular pelvic ultrasonography was used to monitor domi-
nant follicle growth from day 8 of the cycle. Therapy was begun on the day the dominant follicle reached a diameter of 16 mm. A Diasonics DSI-RF sector scanner with a 3.5- to 5-mHz transabdominal transducer (Diasonics, Paris, France) was used for all scans, and mean follicular diameter was calculated from measurements in three planes. The active therapy used was azapropazone (1.2 g daily; four 300-mg capsules), indomethacin (200 mg daily; four 50-mg
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 November 2015. at 18:17 For personal use only. No other uses without permission. . All rights reserved.
NSAIDs AND FOLLICULAR FLUID EICOSANOIDS
237
35 1
30-
25-
20FIG. 2. Comparison of mean follicular fluid PGF2n concentrations and 95% confidence intervals in placebo, azapropazone, indomethacin, and bromfenac groups.
o
15-
Q.
10-
5-
PLACEBO
AZAPROPAZONE
INDOMETHACIN
BROMFENAC
'group difference p=0.007 capsules), or bromfenac (100 mg daily; four 25-mg capsules). All of the capsules (including placebo) were unmarked, and bottles were labeled with a trial number and instructions. A treatment calendar to record any side-effects was given at the time the drugs were dispensed. The date of sterilization was timed so that a large preovulatory follicle would be present. The day before sterilization when the dominant follicle had reached a minimum diameter of 18 mm, 5000 U hCG were given im. hCG was used to eliminate the possibility of any observed effects being due to inhibition of the endogenous gonadotropin surge, as PG inhibition has been shown to be capable of such an effect in animals (15). The operation (and aspiration of follicular fluid) was performed between 17-19.5 h after hCG administration. Sterilization was performed by the minilaparotomy method,
with bilateral tubal ligation. General anesthesia and muscle relaxation were employed in all cases. At the time of sterilization follicular fluid was aspirated from the dominant follicle. Aliquots of fluid were immediately transferred to ice-cold tubes containing indomethacin to inhibit platelet cyclooxygenase activity. The fluid was centrifuged at 900 X g for 10 min. The supernatants were deep frozen to — 20 C. RIA was performed in one batch to avoid interassay variation. Intraassay variation was 8% for the concentrations encountered. Arachidonic acid metabolites were extracted from the follicular fluid essentially as has been described previously for culture medium (16). In brief, the samples were acidified to pH 3 with 15 /uL/niL 8.6 M formic acid and loaded onto preconditioned Sep-Pak C18 columns (Water Associates, Milford, MA). These were washed sequentially with 10 mL each of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 November 2015. at 18:17 For personal use only. No other uses without permission. . All rights reserved.
PRIDDY ET AL.
238
JCE & M • 1990 Vol 71 • No 1
40 i
35-
30 -
25 -
FIG. 3. Comparison of mean follicular fluid LTB4 concentrations and 95% confidence intervals in placebo, azapropazone, indomethacin, and bromfenac groups.
20 -
15 -
10 Ml sags 9
i HP Mi i
5-
i
m 11 I s
m 0
i
J
PLACEBO
AZAPROPAZONE
INDOMETHACIN
BROMFENAC
1
"group difference p=0.005
distilled water and then petroleum spirit (boiling point, 40-60 C) before elution of arachidonic acid metabolites with 5 mL methyl formate into siliconized glass tubes. The methyl formate was evaporated under a stream of nitrogen, and the samples resuspended in 500 nL ethanol and then stored under nitrogen at -70 C. The extraction efficiencies for PGs ranged from 7090% and were independent of the chemical structure, so the mean efficiency (80%) was used to calculate the total PG levels. Likewise, the LTs had extraction efficiencies of 50-60%, which
were independent of the chemical structure; again, the mean efficiency (55%) was used to calculate total levels of LTs. In-house assays were used to measure PGE2, PGF2a, 6-ketoPGFi«, and TXB2 . All antibodies were raised against BSAcoupled eicosanoids, as previously reported (17), and the standards were obtained from Sigma (St. Louis, MO). LTB4 and the peptide LTs were measured using commercial kits (Amersham International, Bucks, United Kingdom). The results are expressed as nanomoles per L follicular fluid, with correction for
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 November 2015. at 18:17 For personal use only. No other uses without permission. . All rights reserved.
NSAIDs AND FOLLICULAR FLUID EICOSANOIDS 1000-i
800o Q.
5 600 5
1 1
• • • Ei
239 PLACEBO AZAPROPAZONE INDOMETHACIN BROMFENAC
400UJ
O 200FIG. 4. Comparison of serum P and E2 concentrations in placebo, azapropazone, indomethacin, and bromfenac groups. SDs are marked.
0
J
40 1 o
1
30-
UJ
O DC UJ
20-
S UJ O
10-
o DC Q.
0J Day of hCG
the extraction efficiencies as stated above. It was not possible to determine individual peptide LT levels, as adequate RIAs to measure each of the peptide LTs separately are not currently available, so these are expressed as a single group. The follicular concentrations of LTs and PGs (except 6-keto-PGFla) were assessed as being log normally distributed after examination of the normal probability plots (18). Therefore, log values were used in the statistical analysis. RIA was performed for estradiol (E2) and progesterone (P) on serum obtained on the day of hCG administration and 7 and 9 days later. All assays were performed in one batch to avoid interassay variation. Intraassay variation for our laboratory was 4% for the E2 assay and 3% for the P assay for the range of values encountered.
Results Of the 41 women recruited to the trial, 11 were in the placebo group, and 10 each in the indomethacin, azapropazone, and bromfenac groups. The mean concentrations of follicular fluid PGE2, PGF2«, 6-keto-PGFla, TXB2, LTB4, and LTC4 (D4 and E4) for each of the four groups were compared using one-
7 days after hCG
9 days after hCG
way analysis of variance. Bromfenac and indomethacin significantly reduced follicular fluid PGE2 (P = 0.011) and PGF2a (P = 0.007) concentrations compared to those after placebo treatment (Figs. 1 and 2). Bromfenac also reduced the LTB4 concentration (P = 0.005) compared to the effect of placebo (Fig. 3). Azapropazone had no effect on follicular fluid eicosanoid concentrations. None of the drugs altered follicular fluid 6-keto-PGFla, TXB2, or LTC4 (D4 and E4) concentrations. Three women complained of nausea, vomiting, and/or dizziness when taking indomethacin. One woman in each of the azapropazone and bromfenac groups complained of nausea. One women felt flushed, and another complained of headache and depression while taking placebo. Differences in mean serum E2 and P concentrations between the groups and over time (i.e. the day of hCG administration and 7 and 9 days later) were assessed using repeated measures analysis of variance. There were no differences in serum E2 and P concentrations among the four groups (Fig. 4). There was a significant decrease in E2 (P < 0.0001) and increase in P (P < 0.0001) in each
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 November 2015. at 18:17 For personal use only. No other uses without permission. . All rights reserved.
PRIDDY ET AL.
240
JCE & M • 1990 Vol 71 • No 1
PLACEBO AZAPROPAZONE INDOMETHACIN BROMFENAC
22 20 18
I1 1
I
(mm)
16 14
UJ
12
N
10O
o Li.
FIG. 5. Comparison of day of cycle and follicular size in placebo, azapropazone, indomethacin, and bromfenac groups. SDs are marked.
864 ,^>'"
2 0
16 14 12 UJ
1 1 1
1
O Q
Tablets commenced
group from the day of hCG administration to day 7 after hCG administration. There was no significant difference in serum E2 or P concentrations between days 7 and 9 after hCG administration in any of the groups.
hCG given
The mean values of subsequent results in the four groups were compared using analysis of variance. There was no significant difference in day of the cycle or size of the follicle when the tablets were commenced, day of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 November 2015. at 18:17 For personal use only. No other uses without permission. . All rights reserved.
NSAIDs AND FOLLICULAR FLUID EICOSANOIDS
241
PLACEBO AZAPROPAZONE INDOMETHACIN
22
o _ FIG. 6. Comparison of time from hCG administration to follicular aspiration (hours) and next menses (days) in placebo, azapropazone, indomethacin, and bromfenac groups. SDs are marked.
BROMFENAC
22
20
20-
18
18-
g g 16
Vol. 71, No. 1 Printed in U.S.A.
The Effect of Prostaglandin Synthetase Inhibitors on Human Preovulatory Follicular Fluid Prostaglandin, Thromboxane, and Leukotriene Concentrations* ALVAN R. PRIDDY, STEPHEN R. KILLICK, MAX ELSTEIN, JULIE MORRIS, MARK SULLIVAN, LATA PATEL, AND MURDO ELDER Departments of Obstetrics and Gynaecology (A.R.P., S.R.K., M.E.) and Statistics (J.M.), University Hospital of South Manchester, Manchester, and the Department of Obstetrics and Gynaecology, Hammersmith Hospital (M.S., L.P., M.E.), London, United Kingdom
ABSTRACT. This study evaluates the eicosanoid concentration in luteinized unruptured follicles (LUFs) on the ovaries of patients who had been treated with inhibitors of prostaglandin synthetase. Indomethacin, bromfenac, or azapropazone (or a placebo) was administered orally to 41 women during the periovulatory period. Follicular development was monitored by serial ultrasound examinations, and the onset of ovulation was regulated by an injection of hCG. Follicular fluid was aspirated during sterilization by minilaparotomy, which was performed
just before the expected time of ovulation. Prostaglandin E2 and PGF2a levels in the fluid were significantly reduced by indomethacin and bromfenac compared to those after placebo treatment. Bromfenac also reduced the follicular fluid leukotriene B4 level. Therefore, the development of luteinized unruptured follicles after treatment with nonsteroidal antiinflammatory drugs appears to be associated with a significant decrease in the synthesis of ovarian eicosanoids. (J Clin Endocrinol Metab 7 1 : 235-242, 1990)
P
indomethacin is not overcome by administering LH at doses that would normally induce ovulation (10). Human follicles have been found to contain PGs, thromboxanes (TXs), and leukotrienes (LTs), although most of the data is from the analysis of follicular fluid from assisted reproduction programs. Human follicular fluid from spontaneous cycles has been found to contain PGE2, PGF2a, 6-keto-PGFla, and TXB 2 (11,12), and the follicular fluid PGF2a concentration increases in the immediate preovulatory period (13). We have previously demonstrated that indomethacin can inhibit ovulation {i.e. induce LUF) in women if given at high doses in the periovulatory period (14). We performed serial ultrasonic scans of follicular development throughout 46 spontaneous cycles in 20 healthy female volunteers, and hCG was given to induce follicular rupture on a particular day. LUFs were seen in 10.7% of untreated cycles, but when NSAIDs were administered over the periovulatory period the incidence of LUF was greatly increased (to 100% with indomethacin; n = 8). The present study was carried out to assess whether NSAIDs (PG synthetase inhibitors) reduce follicular fluid PG, TX, and LT concentrations and to assess the relative effects a new NSAID, bromfenac.
ROSTAGLANDINS (PGs) are involved in mammalian ovulation. The cyclooxygenase inhibitor indomethacin inhibits ovulation in nonprimates such as the rat (1), rabbit (2), pig (3), and sheep (4), and in primates such as the monkey (5). This action is due to PG inhibition, as similar results are obtained using antiserum to either PGF2« (6) or to the cyclooxygenase enzyme itself (7). Administration of PGs overcomes this inhibition. PG concentrations in follicular fluid have been shown to increase in the preovulatory period (8). The LH surge initiates protein synthesis in granulosa cells and cyclooxygenase production via cAMP, so there is a delay from this surge to the increase in the PG content of follicular fluid. Administration of indomethacin at the time of the LH surge of hCG administration reduces the concentrations of follicular fluid PGs and inhibits follicle wall rupture. However, other aspects of the ovulatory response, such as progesterone production, are unaffected, resulting in a luteinized, yet unruptured, follicle (LUF). The action is a local one, as an intrafollicular injection of nonsteroidal antiinflammatory drug (NSAID) prevents ovulation (9). Also, inhibition of ovum release by Received August 4, 1989. Address requests for reprints to: Alvan Priddy, M.R.C.O.G., Department of Obstetrics and Gynaecology, University Hospital of South Manchester, West Didsbury, Manchester, M20 8LR, United Kingdom. * This work was supported by A.H. Robbins Ltd.
Subjects and Methods After obtaining ethical committee approval, 41 caucasion female volunteers were recruited from our waiting list for 235
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 November 2015. at 18:17 For personal use only. No other uses without permission. . All rights reserved.
PRIDDY ET AL.
236
JCE & M • 1990 Vol 71 • No 1
50 i
40 -
*I 51p^
30XAA.
1
m m m
1
FIG. 1. Comparison of mean follicular fluid PGE2 concentration and 95% confidence intervals in placebo, azapropazone, indomethacin, and bromfenac groups.
I
Q.
20"
10-
oPLACEBO
AZAPROPAZONE
INDOMETHACIN
BROMFENAC
*group difference p=0.011 sterilization. Informed written consent was obtained from all volunteers. They were 24-40 yr old, had no gynecological abnormalities, and were not using the oral contraceptive pill. They were randomly assigned to 4 groups to be given a NSAID (indomethacin, azapropazone, or bromfenac) or an indistinguishable placebo orally during the periovulatory period. Randomization was achieved by a computor randomization code. Regular pelvic ultrasonography was used to monitor domi-
nant follicle growth from day 8 of the cycle. Therapy was begun on the day the dominant follicle reached a diameter of 16 mm. A Diasonics DSI-RF sector scanner with a 3.5- to 5-mHz transabdominal transducer (Diasonics, Paris, France) was used for all scans, and mean follicular diameter was calculated from measurements in three planes. The active therapy used was azapropazone (1.2 g daily; four 300-mg capsules), indomethacin (200 mg daily; four 50-mg
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 November 2015. at 18:17 For personal use only. No other uses without permission. . All rights reserved.
NSAIDs AND FOLLICULAR FLUID EICOSANOIDS
237
35 1
30-
25-
20FIG. 2. Comparison of mean follicular fluid PGF2n concentrations and 95% confidence intervals in placebo, azapropazone, indomethacin, and bromfenac groups.
o
15-
Q.
10-
5-
PLACEBO
AZAPROPAZONE
INDOMETHACIN
BROMFENAC
'group difference p=0.007 capsules), or bromfenac (100 mg daily; four 25-mg capsules). All of the capsules (including placebo) were unmarked, and bottles were labeled with a trial number and instructions. A treatment calendar to record any side-effects was given at the time the drugs were dispensed. The date of sterilization was timed so that a large preovulatory follicle would be present. The day before sterilization when the dominant follicle had reached a minimum diameter of 18 mm, 5000 U hCG were given im. hCG was used to eliminate the possibility of any observed effects being due to inhibition of the endogenous gonadotropin surge, as PG inhibition has been shown to be capable of such an effect in animals (15). The operation (and aspiration of follicular fluid) was performed between 17-19.5 h after hCG administration. Sterilization was performed by the minilaparotomy method,
with bilateral tubal ligation. General anesthesia and muscle relaxation were employed in all cases. At the time of sterilization follicular fluid was aspirated from the dominant follicle. Aliquots of fluid were immediately transferred to ice-cold tubes containing indomethacin to inhibit platelet cyclooxygenase activity. The fluid was centrifuged at 900 X g for 10 min. The supernatants were deep frozen to — 20 C. RIA was performed in one batch to avoid interassay variation. Intraassay variation was 8% for the concentrations encountered. Arachidonic acid metabolites were extracted from the follicular fluid essentially as has been described previously for culture medium (16). In brief, the samples were acidified to pH 3 with 15 /uL/niL 8.6 M formic acid and loaded onto preconditioned Sep-Pak C18 columns (Water Associates, Milford, MA). These were washed sequentially with 10 mL each of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 November 2015. at 18:17 For personal use only. No other uses without permission. . All rights reserved.
PRIDDY ET AL.
238
JCE & M • 1990 Vol 71 • No 1
40 i
35-
30 -
25 -
FIG. 3. Comparison of mean follicular fluid LTB4 concentrations and 95% confidence intervals in placebo, azapropazone, indomethacin, and bromfenac groups.
20 -
15 -
10 Ml sags 9
i HP Mi i
5-
i
m 11 I s
m 0
i
J
PLACEBO
AZAPROPAZONE
INDOMETHACIN
BROMFENAC
1
"group difference p=0.005
distilled water and then petroleum spirit (boiling point, 40-60 C) before elution of arachidonic acid metabolites with 5 mL methyl formate into siliconized glass tubes. The methyl formate was evaporated under a stream of nitrogen, and the samples resuspended in 500 nL ethanol and then stored under nitrogen at -70 C. The extraction efficiencies for PGs ranged from 7090% and were independent of the chemical structure, so the mean efficiency (80%) was used to calculate the total PG levels. Likewise, the LTs had extraction efficiencies of 50-60%, which
were independent of the chemical structure; again, the mean efficiency (55%) was used to calculate total levels of LTs. In-house assays were used to measure PGE2, PGF2a, 6-ketoPGFi«, and TXB2 . All antibodies were raised against BSAcoupled eicosanoids, as previously reported (17), and the standards were obtained from Sigma (St. Louis, MO). LTB4 and the peptide LTs were measured using commercial kits (Amersham International, Bucks, United Kingdom). The results are expressed as nanomoles per L follicular fluid, with correction for
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 November 2015. at 18:17 For personal use only. No other uses without permission. . All rights reserved.
NSAIDs AND FOLLICULAR FLUID EICOSANOIDS 1000-i
800o Q.
5 600 5
1 1
• • • Ei
239 PLACEBO AZAPROPAZONE INDOMETHACIN BROMFENAC
400UJ
O 200FIG. 4. Comparison of serum P and E2 concentrations in placebo, azapropazone, indomethacin, and bromfenac groups. SDs are marked.
0
J
40 1 o
1
30-
UJ
O DC UJ
20-
S UJ O
10-
o DC Q.
0J Day of hCG
the extraction efficiencies as stated above. It was not possible to determine individual peptide LT levels, as adequate RIAs to measure each of the peptide LTs separately are not currently available, so these are expressed as a single group. The follicular concentrations of LTs and PGs (except 6-keto-PGFla) were assessed as being log normally distributed after examination of the normal probability plots (18). Therefore, log values were used in the statistical analysis. RIA was performed for estradiol (E2) and progesterone (P) on serum obtained on the day of hCG administration and 7 and 9 days later. All assays were performed in one batch to avoid interassay variation. Intraassay variation for our laboratory was 4% for the E2 assay and 3% for the P assay for the range of values encountered.
Results Of the 41 women recruited to the trial, 11 were in the placebo group, and 10 each in the indomethacin, azapropazone, and bromfenac groups. The mean concentrations of follicular fluid PGE2, PGF2«, 6-keto-PGFla, TXB2, LTB4, and LTC4 (D4 and E4) for each of the four groups were compared using one-
7 days after hCG
9 days after hCG
way analysis of variance. Bromfenac and indomethacin significantly reduced follicular fluid PGE2 (P = 0.011) and PGF2a (P = 0.007) concentrations compared to those after placebo treatment (Figs. 1 and 2). Bromfenac also reduced the LTB4 concentration (P = 0.005) compared to the effect of placebo (Fig. 3). Azapropazone had no effect on follicular fluid eicosanoid concentrations. None of the drugs altered follicular fluid 6-keto-PGFla, TXB2, or LTC4 (D4 and E4) concentrations. Three women complained of nausea, vomiting, and/or dizziness when taking indomethacin. One woman in each of the azapropazone and bromfenac groups complained of nausea. One women felt flushed, and another complained of headache and depression while taking placebo. Differences in mean serum E2 and P concentrations between the groups and over time (i.e. the day of hCG administration and 7 and 9 days later) were assessed using repeated measures analysis of variance. There were no differences in serum E2 and P concentrations among the four groups (Fig. 4). There was a significant decrease in E2 (P < 0.0001) and increase in P (P < 0.0001) in each
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 November 2015. at 18:17 For personal use only. No other uses without permission. . All rights reserved.
PRIDDY ET AL.
240
JCE & M • 1990 Vol 71 • No 1
PLACEBO AZAPROPAZONE INDOMETHACIN BROMFENAC
22 20 18
I1 1
I
(mm)
16 14
UJ
12
N
10O
o Li.
FIG. 5. Comparison of day of cycle and follicular size in placebo, azapropazone, indomethacin, and bromfenac groups. SDs are marked.
864 ,^>'"
2 0
16 14 12 UJ
1 1 1
1
O Q
Tablets commenced
group from the day of hCG administration to day 7 after hCG administration. There was no significant difference in serum E2 or P concentrations between days 7 and 9 after hCG administration in any of the groups.
hCG given
The mean values of subsequent results in the four groups were compared using analysis of variance. There was no significant difference in day of the cycle or size of the follicle when the tablets were commenced, day of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 23 November 2015. at 18:17 For personal use only. No other uses without permission. . All rights reserved.
NSAIDs AND FOLLICULAR FLUID EICOSANOIDS
241
PLACEBO AZAPROPAZONE INDOMETHACIN
22
o _ FIG. 6. Comparison of time from hCG administration to follicular aspiration (hours) and next menses (days) in placebo, azapropazone, indomethacin, and bromfenac groups. SDs are marked.
BROMFENAC
22
20
20-
18
18-
g g 16
