CYP2D6 poor metabolizers produce less endoxifen (32), thus providing a possible explanation for the genotype differ- ence in effect. CC is also 4-hydroxylated ...
OVULATION INDUCTION Evaluation of the relationship between plasma concentrations of en- and zuclomiphene and induction of ovulation in anovulatory women being treated with clomiphene citrate Cyrus Ghobadi, M.B.Ch.B.,a,b Saad Amer, M.D.,b Hany Lashen, M.D.,b Martin S. Lennard, Ph.D.,a William L. Ledger, D.Phil.,b and Amin Rostami-Hodjegan, Ph.D.a a
Academic Unit of Clinical Pharmacology; and b Academic Unit of Reproductive and Developmental Medicine, University of Sheffield, Sheffield, United Kingdom
Objective: To investigate the relationship between the plasma concentrations of clomiphene citrate (CC) isomers zu- (Zu) and enclomiphene (En), and ovulation outcome. Design: Prospective, cohort study. Setting: Reproductive medicine and fertility center in a university teaching hospital, United Kingdom. Patient(s): Forty-two women with World Health Organization type 2 infertility. Intervention(s): The clinical and biochemical features of patients who were about to start CC for induction of ovulation were recorded. Plasma concentration of Zu and En were monitored at three points (days 2, 8, and 21) throughout the treatment cycle(s). Main Outcome Measure(s): Ovulation. Result(s): Thirty-nine patients completed the study. Both En and Zu accumulated throughout treatment. Among the 36 responders, there was no statistically significant relationship between the clinical and biochemical characteristics of the patients, En or Zu concentrations, and the dose required to induce ovulation. Moreover, the Zu and En concentrations were not different in the three patients who failed to respond. Conclusion: The concentrations of En and Zu in plasma, on their own or in combination with other covariates (e.g., weight, body mass index, free androgen index), are not a predictor of the ovulation response to CC or of the dose requirement. Further studies are needed to explore the role of additional covariates, including the presence of active metabolites, and the balance of the effects of En and Zu. (Fertil Steril 2009;91:1135–40. 2009 by American Society for Reproductive Medicine.) Key Words: Clomiphene citrate, concentration–response, clomiphene isomers, ovulatory response, infertility treatment, normogonadotropic anovulatory infertility
Clomiphene citrate (CC) is the best initial infertility treatment for the large majority of women with absent or irregular ovulation. These patients are classified as having normogonadotropic anovulatory infertility (World Health Organization [WHO] type 2) (1, 2). Although new treatments, for example metformin (3), have been gaining popularity in clinical practice, and large trials are in progress to evaluate the efficacy of these treatments, none of the studies have demonstrated any
Received December 25, 2007; revised January 12, 2008; accepted January 14, 2008; published online March 18, 2008. C.G. has nothing to disclose. S.A. has nothing to disclose. H.L. has nothing to disclose. M.S.L. has nothing to disclose. W.L.L. has nothing to disclose. A.R.-H. has nothing to disclose. C.G. is a Ph.D. candidate, and the present work constitutes part of his thesis. Reprint requests: Amin Rostami-Hodjegan, Academic Unit of Clinical Pharmacology, Floor M, The Royal Hallamshire Hospital, Sheffield S10 2JF, United Kingdom (FAX: 44-114-2711863; E-mail: a.rostami@ sheffield.ac.uk).
0015-0282/09/$36.00 doi:10.1016/j.fertnstert.2008.01.058
evidence for the superiority of alternative treatments (4–6). Thus, the use of CC remains a well-validated first-line therapy in patients with WHO type 2 infertility (3, 5, 7–11). Most studies have reported ovulation rates of 60%–85% and pregnancy rates of 30%–40% with CC (7). There is a significant variation in the dose required (50–150 mg) to induce ovulation (12) with CC, which is not predictable a priori. Thus, several months of treatment may be required to achieve the optimal dose, and this poses a problem in use of CC. Recent reports have indicated that a number of individual patient characteristics, such as age, body weight, body mass index (BMI), free androgen index (FAI), and menstrual cycle history contribute to the success or failure of treatment with CC (13–16). However, little attention has been paid to the interindividual variability in CC metabolism and its potential contribution to the therapeutic response (17–20). In theory, variation in biotransformation of CC will propagate to its concentrations in plasma, leading to differences in the
Fertility and Sterility Vol. 91, No. 4, April 2009 Copyright ª2009 American Society for Reproductive Medicine, Published by Elsevier Inc.
1135
latter for any given dose. Consequently, monitoring plasma concentration of CC might offer an independent predictor of outcome at an early stage of treatment and give an indication of the desired dose in each individual (21). CC is administered as a mixture of two geometric isomers, enclomiphene (En) and zuclomiphene (Zu) citrate in the ratio 62%:38% (22). The aim of the present study was to investigate the relationship between plasma concentrations of En and Zu, on their own or in combination with other patient covariates, as predictors of therapeutic outcome in a group of infertile women undergoing CC treatment. MATERIALS AND METHODS Patients This study was conducted over 2 years at the Centre for Reproductive Medicine and Fertility, Jessop Wing, Royal Hallamshire Hospital, Sheffield, United Kingdom. All patients who presented with infertility problems were assessed with a standardized evaluation, including testosterone (T), serum hormone-binding globulin (SHBG), and FAI. Subsequent evaluation was conducted to identify anovular patients and rule out other causes of infertility, such as tubal problems, hyperprolactinemia, hypo/hyperthyroidism, premature ovarian failure, male factor, and unexplained infertility. Anovular patients with relevant anatomic and adnexal pathologies, or with serum prolactin (PRL) and thyroid-stimulating hormone (TSH) concentrations out of the normal range (PRL: 59– 619 mIU/L; TSH: 0.35–4.5 mIU/L), or with elevated serum concentrations of FSH in the early follicular phase (defined as day 2–4 FSH >23 IU/L) (23) were excluded from the study. Patients with previous exposure to ovulation induction agents or combination therapy were also excluded. The study was approved by the South Sheffield Research Ethics Committee. Informed consent was obtained from all participants. Treatment Protocol Subjects who met the inclusion criteria received an initial dose of 50 mg/d CC on days 2–6 according to the current stepwise protocol for CC therapy. Ovulation was assumed to have occurred if the midluteal phase serum progesterone (P) concentration was >20 ng/mL. If patients failed to respond, the dose of CC was increased to 100 mg in the next cycle, and then to 150 mg if there was no response to 100 mg. The drug was discontinued if patients failed to ovulate after three consecutive cycles at the highest dose. All patients were instructed to take the drug once daily on days 2–6 of their menstrual cycle. They were also asked to record the time of dosing. Blood samples (10 mL) were taken from patients on days 2 (between 1 and 24 hours after the dose, the time being recorded), 8, and 21 in each cycle of treatment, and plasma samples were frozen and stored (20 C) until assayed. 1136
Ghobadi et al.
FIGURE 1 Response to CC at different doses. R50 ¼ responders to 50 mg; NR50 ¼ nonresponders to 50 mg; R100 ¼ responders to 100 mg; NR100 ¼ nonresponders to 100 mg; R150 ¼ responders to 150 mg; NR150 ¼ nonresponders to 150 mg.
Ghobadi. Clomiphene concentrations and response. Fertil Steril 2009.
Analysis of Hormonal, Zu, and En Concentrations Serum concentrations of LH, FSH, TSH, PRL, and midluteal P were measured as part of routine clinical practice with use of well-established and validated assays at the Department of Clinical Chemistry, Royal Hallamshire Hospital. Details of these assays have been published previously (24). These data had also been used to establish the cause of infertility. Serum concentration of Zu and En were measured by liquid chromatography–mass spectrometry, as described by Crewe et al. (25). The limits of determination were 35 pg/mL and 7 pg/mL for Zu and En, respectively. Corresponding withinday coefficients of variation ranged from 2.1% to 7.2%.
Statistical Analysis Statistical analysis was performed with the SPSS statistical software for Windows (version 12; SPSS, Chicago, IL). Patients were divided into three groups according to their final dose of CC. Plasma concentrations of Zu and En and
Clomiphene concentrations and response
Vol. 91, No. 4, April 2009
FIGURE 2 Plasma concentrations of En (top) and Zu (middle) and their ratio (bottom) during the three cycles of treatment with CC. Values are expressed as mean SE; the vertical broken line on day 21 (D21) indicates the end of the cycle when patients were assessed for ovulation. A ¼ responders and nonresponders to 50 mg; B ¼ responders and nonresponders to 100 mg; C ¼ responders and nonresponders to 150 mg.
50 mg
100 mg
21 21 D
D
8
D
D
8
D
21
8 D
2 D 2 D 2
D
21
D
21
8
2 D
21
D 8 D
D
50 mg
D
D
D
D
D
100 mg
21
0 D
0
8
0
D
50
2
50
D 21
50
D 8
100
D 2
100
D 21
100
D 8
150
D 2
150
21
150
8
200
50 mg
D
2
D
D
8 D
D
D
D
2 D
D
D
D
200
2
200
2
0 D
0
2
0
D
10
21
10
D
10
21
20
D
20
21
20
2
30
21
30
8
30
21
40
8
40
8
2 D
21 D
D
D
40
D
0
8
0
8
0
D 21
2 D 8
2 D 2
4
2
D 21
4
D 8
6
4
D 2
6
8
8
6
Ovulated Did not ovulate
C
10
8
2
Zu (ng/ml) Zu/En ratio
10
8
2
En (ng/ml)
10
B
D
A
150 mg
Observation time (day)
Ghobadi. Clomiphene concentrations and response. Fertil Steril 2009.
the Zu/En ratio were plotted against time points for all treatment cycles. We used simple imputation for values below the limit of detection (26, 27). Thus all values below the quantification limit (BQL) of the assay were set to half the quantification limit. The c2 test was used to compare sets of data before and after exclusion of BQL samples to detect any possible bias that may have been introduced by an unequal distribution of BQL values in the different group of patients. The cumulative pattern of Zu and En concentrations and the Zu/En ratio were investigated throughout treatment on days 2, 8, and 21 with repeated-measures analysis of variance (ANOVA). Differences in the concentration of Zu and En between groups of ovulatory and non-ovulatory patients were also studied within treatment cycles with one-way ANOVA. Post hoc comparisons were performed with Bonferroni correction for multiple comparisons. Fertility and Sterility
In addition, observations from the first cycle of CC therapy were assessed after dividing patients into four groups according to the final outcome (i.e., responders to 50 mg, 100 mg, and 150 mg and resistant to therapy). Subsequently, patient characteristics including age, BMI, LH concentrations, FSH concentrations, LH/FSH ratio, SHBG concentrations, FAI, and the Zu/En concentration ratio were compared between all groups (one-way ANOVA) with respect to therapeutic outcome. Age, BMI, FAI, and cycle history were used to estimate the chance of ovulation for each patient on the basis of a published nomogram (16). Logistic regression with a backward stepwise exclusion procedure was used to determine variables influencing ovulation. Three outcome measures were the subject of logistic regression analysis, namely [1] response or lack of response to 50 mg dose of CC, [2] response or lack of response to CC 1137
FIGURE 3
5 4 Eventual response: 3
a
ovulated on 50mg ovulated on 100mg ovulated on 150mg Resistant to treatment
2 1
a
D
D
D
8
21
0 2
LN Zu/En in the first cycle of treatment
The natural logarithm (LN) (mean SE) of the Zu/En ratio in plasma, in the first cycle of treatment. Patients are grouped retrospectively according to the final outcome of their therapy. aP< .05.
Time (day) Ghobadi. Clomiphene concentrations and response. Fertil Steril 2009.
dose up to 100 mg, and [3] response or lack of response to CC doses up 150 mg. The area under the curve of receiver operating characteristics (ROC) was used to assess the diagnostic accuracy of any predictive covariate model. Statistical significance was assumed when the P value was < .05. All values are expressed as mean SE. RESULTS Of a total 42 patients who participated in this study, 3 did not fully comply with the protocol and were therefore excluded from the analysis. Of the remaining 39, 23 (59%) ovulated
with 50 mg CC, 11 (28%) with 100 mg, 2 (5%) with 150 mg, and 3 (8%) did not ovulate and were considered resistant to CC (Fig. 1). Figure 2 shows the cumulative pattern and the mean plasma concentration of Zu and En in three cycles of treatment in the group of ovulatory and non-ovulatory patients. Both Zu and En showed an overall cumulative increase throughout the course of treatment (P
