REPRODUCTIVE ENDOCRINOLOGY

Transcription

1 REPRODUCTIVE ENDOCRINOLOGY FERTILITY AND STERILITY VOL. 73, NO. 1, JANUARY 2000 Copyright 1999 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A. Neuroendocrine response to an intravenous L-tryptophan challenge in women with premenstrual syndrome Natalie Rasgon, M.D., Ph.D.,* Michael McGuire, M.D.,* Sohrab Tanavoli,* Lynn Fairbanks, Ph.D.,* and Andrea Rapkin, M.D. University of California, Los Angeles School of Medicine, Los Angeles, California Received April 5, 1999; revised and accepted July 30, Supported by United States Public Health Service Grant No. M01RR Presented at the Society for Gynecologic Investigation, Atlanta, Georgia, United States, March 10 13, Reprint requests: Andrea Rapkin, M.D., Department of Obstetrics and Gynecology, UCLA School of Medicine, Le Conte Avenue, Room CHS, Los Angeles, California (FAX: ; * Department of Psychiatry and Biobehavioral Sciences. Department of Obstetrics and Gynecology /99/$20.00 PII S (99) Objective: To evaluate the neuroendocrine responses to an intravenous L-tryptophan challenge across the menstrual cycle in women with premenstrual syndrome (PMS) and controls. Design: Controlled clinical study. Setting: The Clinical Research Center of an academic research environment. Patient(s): Women with PMS and healthy volunteers. Intervention(s): An intravenous L-tryptophan challenge was administered two times a week during 1 month to five subjects with prospectively documented PMS and five age- and body mass-matched controls. Main Outcome Measure(s): Whole-blood serotonin, cortisol, and prolactin levels were assessed at the baseline and at 30, 50, 60, 70, 80, and 90 minutes after the challenge. Result(s): Whole-blood serotonin response to the L-tryptophan challenge was blunted in the luteal phase of the menstrual cycle in subjects with PMS compared with controls. Cortisol levels differed between groups and cycle phases only at the baseline, with higher baseline cortisol levels during the luteal phase in women with PMS, whereas baseline and postchallenge prolactin levels did not differ between groups. Conclusion(s): The present results support previously reported findings of alterations in tryptophan handling in women with PMS. The elevated baseline luteal phase cortisol concentrations in subjects with PMS warrants further investigation. (Fertil Steril 2000;73: by American Society for Reproductive Medicine.) Key Words: L-tryptophan, serotonin, premenstrual syndrome Serotonin is an important neurotransmitter modulating mood, behavior, appetite, and sexual function. Abnormalities of serotonergic function have been associated with affective disorders and premenstrual syndrome (PMS). Serotonin is also a putative mediator in the hypothalamic-pituitary adrenal axis modulating the output of cortisol and prolactin (1). We previously showed that in PMS there is a serotonergic deficiency as manifested by lower luteal phase whole-blood serotonin (5- HT) levels and abnormal luteal phase response to oral tryptophan loading test (2, 3). Serotonergic dysfunction in PMS also has been postulated on the basis of findings of decreased platelet uptake of serotonin (4), blunted response to 5-HT agonists (5), and consistent therapeutic efficacy of selective serotonergic reuptake inhibitors (SSRIs) in the treatment of this disorder (6). Pharmacological challenge paradigms have been used to examine the role of 5-HT functioning in the pathogenesis of depression (7 13) and PMS (3, 5, 14). This approach is based on the ability of agents that modulate 5-HT function to elicit neuroendocrine responses such as secretion of prolactin (PRL) and cortisol. Intravenous (IV) infusion of L-tryptophan (LTP) has been used as a pharmacological probe to assess the role of 5-HT in affective disorders. Most studies have demonstrated a differential prolactin and cortisol response between depressed patients and healthy control subjects (8 13). Administration of IV LTP to women with premenstrual depression revealed a blunted prolactin response in the luteal phase in both depressed and asymptomatic women (14). In this study we further evaluated the activity of the serotonergic pathway across the men- 144

2 strual cycle in women with PMS and determined whether administration of the 5-HT precursor LTP led to an alteration in whole-blood serotonin levels. We also wanted to determine whether women with PMS would manifest changes in cortisol or prolactin concentrations, which may be different from those found in controls either at baseline or after IV administration of the serotonin precursor LTP. MATERIALS AND METHODS Subject Selection Subjects were recruited through advertisements in local newspapers and solicitation of University of California, Los Angeles students. The study was approved by the Human Subjects Protection Committee at the University of California, Los Angeles. All subjects gave informed consent for their participation. Subjects were evaluated for 2 months before acceptance in the study. During evaluation, subjects were required to complete a daily PMS prospective symptom diary. The premenstrual diary consisted of 11 of the more commonly experienced PMS symptoms including depression, anxiety, mood swings, irritability or anger, headaches, breast pain, low energy, edema, food cravings, avoidance of social activity, impaired relationships, and diminished work performance. Each symptom was rated on a scale of 1 (not at all) to 6 (extreme). In the follicular phase of their second menstrual cycle, all subjects underwent a Structured Clinical Interview (SCID- NP) for Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) (15) to confirm the absence of significant current or recent (past 2 years) major psychiatric illness, including substance and alcohol abuse. Eligibility criteria for PMS subjects included age years, history of regular menses with moderate to severe premenstrual symptoms for at least the previous 6 months, and negative medical history other than PMS. Accepted subjects had to meet the DSM-IV characteristics of premenstrual dysphoric disorder (15). All women were screened for the absence of significant medical illness (current or in the past 2 years) through history and physical examination. The body mass index (BMI) for all subjects was calculated as the weight in kilograms divided by the square of the height in meters. Women with a BMI exceeding 25 were excluded. None of the women recruited into the study were exposed to psychoactive medications, hormonal preparations (including oral contraceptives and thyroid replacement), mineral supplements and vitamins within the past 3 months. All participants were required to use nonsteroidal contraception to prevent pregnancy during the course of the study. Subjects with PMS who failed to demonstrate moderate to severe symptom scores on the daily diary with at least a 50% increase in a minimum of four of the psychological symptoms and two of the physical symptoms during the luteal phase compared with the follicular phase were excluded. Once accepted into the study, all subjects were required to continue to complete the PMS symptom diary for the ensuing month of the study. Eligibility criteria for control subjects were the same except that premenstrual symptoms were absent. Seventyfive women were screened for the study: 28 women with PMS and 47 control subjects. Of 28 women with PMS, 18 were excluded for one or more of the following reasons: fear of venipuncture, lack of time, unwillingness to comply with diet and alcohol restrictions for 6-week periods, premenstrual exacerbation of an underlying psychiatric disorder, comorbidity with substance abuse, and anxiety disorders. Three subjects moved out of the area at the time of the beginning of the challenge phase, one patient developed fever at the time of the first Clinical Research Center (CRC) visit. Of 47 control subjects 22 rejected the study on the basis of the time constraints and fear of venipuncture, irregular menses, psychiatric condition other than PMS, or unwillingness to use nonsteroidal contraception. One woman had an anovulatory cycle. After exclusions, 6 women with PMS and 7 control subjects entered the challenge phase of the study. One PMS subject and 2 controls withdrew from the study after the first visit secondary to lack of interest, thus 5 subjects in each group completed the study. Procedure The length of the challenge study was 1 month. All subjects were required to abstain from alcohol and to follow a tyramine-free diet for 2 weeks before the challenge phase and throughout the study. Subjects presented to the CRC at the UCLA Medical Center two times a week for 4 consecutive weeks. Subjects were also encouraged to keep their visits on the same days of the week (e.g., Thursday Sunday or Tuesday Friday) to eliminate fluctuations in the schedule. During the challenge month, subjects were asked to document ovulation using a urinary LH-kit (Ovu-kit; Quidel Corp., San Diego, CA). The day of LH surge was considered as day 0. Follicular phase (inclusive of middle to late follicular phases) was considered as ovulation days ( 12) to ( 5), ovulatory phase was calculated as ovulation days ( 4) to ( 3), and luteal phase (inclusive of middle to late lute phases) was calculated as ovulation days ( 7) to ( 14). The phase of the menstrual cycle in which blood was obtained was confirmed retrospectively by the date of onset of the next menstrual period, and ovulation was reconfirmed by the luteal phase serum progesterone levels. Neuroendocrine Challenge LTP challenge was conducted as described by Price et al (8). The LTP infusions were prepared by dissolving 8.4 g of L-tryptophan in 600 ml of 0.45% saline solution, with 50% sodium hydroxide added to bring the solution to a ph level of 7.4. Each 600-mL aliquot was sterilized by a passage FERTILITY & STERILITY 145

3 through a 0.22-mm filter (Millipore Corp., Bedford, MA) and was tested for pyrogenicity and sterility before use (8). On the test day, all subjects arrived at the CRC at 7:00 A.M. after an overnight fast and remained fasting until the end of the test. Beginning at 7:00 A.M. subjects rested on their beds supine with their heads slightly elevated, until the completion of the test (approximately 4 hours later). They were allowed to ambulate to the bathroom, but not permitted to sleep. At 7:30 A.M., an IV catheter was placed in the forearm vein and was kept open with a flush of heparin. At 9:00 A.M. subjects received an IV solution of LTP (7.0 g of tryptophan diluted in 500 ml of 0.47% normal saline) given at a constant rate for a 20-minute period. Samples were obtained before the start of the tryptophan infusion (T 15 and T 0), and at 30, 50, 60, 70, 80, and 90 minutes after the start of the infusion. Serum and whole-blood samples were frozen at 70 C until assayed. Whole-Blood Serotonin Assay Whole-blood serotonin levels were determined by a highpressure liquid chromatography method with fluorescent detection as described by Brammer (16). Serotonin was resolved from tyrosine and tryptophan on a C-18 analytical column ( mm, 5 m particle size) and detected by fluorescence at 283 nm excitation and 330 emission. Quantitation was achieved by reference to a standard line derived from external standards treated the same as samples, and results were expressed as ng analyte per ml whole blood (16). All assays were conducted in duplicates taken as the actual value. The interassay coefficient of variation was 6%, the intra-assay variability was 4%. Cortisol Cortisol was measured with use of the Cortisol I-125 radioimmunoassay method by Diagnostic System Laboratories (Webster, TX). This assay has a cross-reactivity of 33.3% and 9.3% with prednisolone and corticosterone, respectively. The intraassay and interassay coefficients of variation of replicate samples were 8.3% and 9.8%, respectively. Prolactin Prolactin was measured by using the immunoradiometric assay by Diagnostic System Laboratories described in Miles et al. (17). The intraassay and interassay coefficients of variation of replicate samples were 3.6% and 8.0%, respectively. Statistical Analysis Diary data were evaluated by nonparametric statistics using the Mann-Whitney U test (18). The neuroendocrine (whole-blood serotonin, cortisol, and prolactin) responses to LTP challenge were assessed during the follicular, ovulatory, and luteal phases of the menstrual cycle. The change in responses (delta max) for each variable was calculated by subtracting the mean baseline value from the peak value observed at one of the 30- to 90-minute time points. Differences between PMS and control subjects in baseline and delta max scores were evaluated with use of multivariate analysis of variance with diagnosis as a between-subject factor and cycle phase as a within-subject factor. Missing values for one cycle phase were replaced by the group mean for two subjects. All results are reported as significant when P.05. RESULTS Subject Characteristics The age range was from years for both groups with mean ( SD) age of 24 0 for subjects with PMS and 27 4 for controls. All study subjects were unmarried and nulliparous. There were no statistically significant differences in the mean ages or BMI between groups. Table 1 shows the mean symptom scores for PMS and control subjects during the follicular (ovulation days [ 12] to [ 5]) and luteal (ovulation days [ 7] to [ 14]) and luteal phases of the cycle. Luteal phase symptom scores in women with PMS were significantly greater than controls for most symptoms (Table 1). None of these comparisons differed significantly in the follicular phase. Whole-Blood Serotonin Levels The baseline (t 15, t 0) levels and delta max 5-HT values (peak minus baseline) for subjects with PMS and for controls in follicular, ovulatory, and luteal phases are presented in Table 2. Responses to LTP challenge were influenced by both diagnosis and cycle phase. There was a significant interaction between diagnosis and cycle phase (F 6.72, df 2,7, P.05) with the subjects with PMS showing a markedly attenuated response to challenge in the luteal phase compared with controls. Delta max 5-HT responses were more robust during the follicular phase in both groups. Cortisol Mean baseline values of cortisol and delta max responses to IV LTP in each phase of the menstrual cycle are shown in Table 3. Baseline cortisol levels were similar in subjects with PMS and in control subjects during the follicular phase but significantly higher in subjects with PMS in the luteal phase of the menstrual cycle. This difference was reflected in a significant within-subjects effect for cycle phase (F 6.65, df 2,7, P.05) and a significant diagnosis by cyclephase interaction (F 5.21, df 2,7, P.05) (Table 3). There was considerable variability in response among the subjects with PMS and none of these postchallenge differences in cortisol were statistically significant. Log transformation of the data to reduce the dependency between means and variances did not alter the results. Prolactin There were no significant effects of cycle phase or diagnosis on baseline levels of prolactin (Table 3). There was a tendency for the prolactin response to challenge to be greater in the ovulatory and luteal phase than during the follicular phase (cycle phase F 3.49, df 2,16, P.055). Subjects with PMS exhibited more variation than the controls in response to challenge, but the average peak change from 146 Rasgon et al. IV L-tryptophan challenge in PMS Vol. 73, No. 1, January 2000

4 TABLE 1 Daily diary symptom scores in women with PMS and controls in the follicular and luteal phases of the cycle. Study group PMS Control Symptom Follicular Luteal Follicular Luteal Avoid social activity ( ) ( )* ( ) ( ) Decreased interest ( ) ( )* ( ) ( ) Edema/weight gain ( ) ( )* ( ) ( ) Depression ( ) ( ) ( ) ( ) Anxiety ( ) ( ) ( ) ( ) Mood swings ( ) ( ) ( ) ( ) Irritability ( ) ( ) ( ) ( ) Increased appetite/cravings ( ) ( )* ( ) ( ) Low energy/fatigue ( ) ( ) ( ) ( ) Headaches ( ) ( )* ( ) ( ) Breast pain ( ) ( ) ( ) ( ) Sleep disturbance ( ) ( ) ( ) ( ) Difficulty concentrating ( ) ( ) ( ) ( ) Note: Values are means SD (range). * P.05 (PMS vs. control group). P.01 (PMS vs. control group). Rasgon. Neuroendocrine response to L-tryptophan. Fertil Steril baseline did not differ between the two groups. Log transformation of the data to reduce the variance did not alter the results. DISCUSSION The main objective of this study was to evaluate the 5-HT response to an IV LTP challenge across the menstrual cycle in women with and without PMS. The second objective was to assess whether the difference in 5-HT response to this challenge could be correlated with the postchallenge changes in prolactin and cortisol concentrations. Our observations indicated that there was a difference in LTP handling in subjects with PMS compared with controls only in the luteal phase. The response to LTP handling in the follicular phase was similar in women with and without PMS, because in both groups, there was an increase in serotonin levels. However, in the luteal phase the response to LTP handling was statis- TABLE 2 Whole-blood serotonin responses to IV L-tryptophan challenge across the menstrual cycle in women with PMS and controls. Group Baseline Serotonin response (ng/ml) Delta max PMS Follicular Ovulatory Luteal * 20.0 Control Follicular Ovulatory Luteal Note: Values are means SD. * P.05 (PMS subjects vs. controls). Rasgon. Neuroendocrine response to L-tryptophan. Fertil Steril TABLE 3 Cortisol and prolactin responses to IV L-tryptophan challenge across the menstrual cycle in women with PMS and controls. Group Cortisol level ( g/dl) Baseline Delta max Prolactin level (ng/ml) Baseline Delta max PMS Follicular Ovulatory Luteal * Control Follicular Ovulatory Luteal * Note: Values are means SD. * P.05 (PMS subjects vs. controls). Rasgon. Neuroendocrine response to L-tryptophan. Fertil Steril FERTILITY & STERILITY 147

5 tically significantly different in control women compared with women with PMS. In control women there was a rise in 5-HT levels similar to that observed in the follicular phase in women with and without PMS, whereas the response was markedly blunted in the luteal phase in women with PMS. The changes in subjects with PMS were only observed in the luteal phase. This indicates that a factor in the luteal phase in subjects with PMS must differ from that of controls. Our study confirms previous findings using an oral tryptophan loading test in which 5-HT concentrations also declined in the luteal phase in subjects with PMS (3). In the oral loading test, one could argue that there were variabilities in absorption that may have accounted for the differences in handling between the follicular and luteal phase and between groups because progesterone is known to decrease gastrointestinal motility. However, the observation that iv loading also shows this alteration firmly establishes that there is a difference in handling of tryptophan in subjects with PMS in the luteal phase compared with follicular phase and with controls. The differences in the disposition of LTP were not reflected in changes in serotonergic control of hypothalamic, pituitary, or adrenal functioning because there were no differences between groups or across time in delta max cortisol or prolactin responses. Another interesting finding in the current study are the elevated baseline cortisol levels in the luteal phase that were observed only in subjects with PMS. In women with PMS, differences in baseline cortisol levels have not been previously noted (5, 14, 19). It is difficult to determine whether the high baseline cortisol levels are a result of abnormal handling of the tryptophan load because serotonin does modulate ACTH release or due to increased stress or anxiety in these subjects. Enhanced cortisol response to corticotropin-releasing hormone has been observed in women with PMS (19). Similar to the findings of Bancroft et al. (14), we found no significant diagnosis-related changes either in the baseline PRL levels or in PRL responses to LTP. Furthermore, in agreement with Su et al. (5) who used a 5-HT agonist m-chlorophenyl piperazine as a neuroendocrine challenge agent, we have identified a trend for greater LTP-induced PRL secretion in the ovulatory and luteal phases of the menstrual cycle that may be related to the stimulatory effect of estradiol on 5-HT secretion. Intravenous L-tryptophan loading has been used to examine 5-HT functioning in depression (7 13). Most studies have demonstrated a differential response between depressed patients and healthy controls as manifested by blunted PRL responses (7 13), although both blunted (10) and exaggerated (7) cortisol responses to precursor challenge have been noted in patients with affective disorders. However, there are profound differences between PMS and affective disorders, and PMS is not considered a subtype of depressive disorder (15). Biological markers that were abnormal in individuals with affective disorders, such as platelet monoamine oxidase B (20) and dexamethasone suppression test (21), have not been altered in PMS (22, 23). Further supporting this divergence in the disorders is the fact that although SSRIs have therapeutic efficacy in both the treatment of major depression and PMS, the time for onset of action is 4 6 weeks in affective disorder, whereas symptomatic response to SSRIs occurs within days in women with PMS. Because there were no between-group differences in postchallenge PRL or cortisol levels, the current study further supports the concept that the pathogenesis of affective disorders and PMS differ, although 5-HT dysfunction is likely to be characteristic of both disorders. It would be interesting to attempt to determine the enzymatic differences or the mechanism by which the L-tryptophan handling is different in subjects with PMS compared with controls in the luteal phase and whether this mechanism may be in part responsible for the symptoms of premenstrual syndrome. In summary, this pilot study contributes to the evidence of menstrual cycle phase-related changes in serotonergic activity in women with PMS. Further studies are in progress to evaluate in a larger sample of subjects whether the wholeblood serotonin and prolactin changes in response to IV LTP in the early, middle, and late follicular and luteal phases of the menstrual cycle are different in women with PMS compared with controls and to delineate more specific interactions between ovarian sex steroids and serotonin. Attempts to replicate and expand on the findings of elevated baseline cortisol levels in subjects with PMS and its relationship to symptoms of PMS are also warranted. References 1. Lesch KP, Müller U, Rupprecht R, Kruse K, Schulte HM. Endocrine responses to growth hormone-releasing hormone, thyrotropin-releasing hormone and corticotropin-releasing hormone in depression. Acta Psychiatr Scand 1989;79: Rapkin AJ, Edelmuth E, Chang LC, Reading AE, McGuire MT, Su TP. Whole-blood serotonin in premenstrual syndrome. Obstet Gynecol 1987;70: Rapkin AJ, Chuong LC, Reading A, McGuire MT. Tryptophan loading test in premenstrual syndrome. Obstet Gynecol 1989;10: Taylor DL, Mathew RJ, Beng TH. Serotonin levels and platelet uptake during premenstrual tension. Neuropsychobiology 1984;12: Su TP, Schmidt PJ, Danaceau M, Murphy DL, Rubinow DR. Effect of menstrual cycle phase on neuroendocrine and behavioral responses to the serotonin agonist m-chlorophenylpiperazine in women with premenstrual syndrome and controls. J Clin Endocrinol Metab 1997;82: Steiner M, Steinberg S, Stewart D, Carter D, Berger C, Reid R, et al. Fluoxetine in the treatment of the premenstrual dysphoria. N Engl J Med 1997;332: Meltzer HY. Role of serotonin in depression. Ann NY Acad Sci 1990;600: Price LH, Charney DS, Delgado PL, Heninger GR. Serotonin function and depression: neuroendocrine and mood responses to intravenous L-tryptophan in depressed patients and healthy comparison subjects. Am J Psychiatry 1991;148:11, Delgado PL, Charney DS, Price LH, Ahajanian GK, Landis H, Heninger GR. Serotonin function and the mechanism of antidepressantinduced remission by rapid depletion of plasma tryptophan. Arch Gen Psychiatry 1990;47: Cowen PJ, Charig EM. Neuroendocrine responses to intravenous tryptophan in major depression. Arch Gen Psychiatry 1987;44: Heninger GR, Charney DS, Sternberg DE. Serotonergic function in 148 Rasgon et al. IV L-tryptophan challenge in PMS Vol. 73, No. 1, January 2000

6 depression: prolactin response to intravenous tryptophan in depressed patients and healthy subjects. Arch Gen Psychiatry 1984;41: Price LH, Malison RT, McDougle CJ, Pelton GH, Heninger GR. The neurobiology of tryptophan depletion in depression: effects of intravenous tryptophan infusion. Biol Psychiatry 1998;43: Yatham L, Steiner M. Neuroendocrine probes of serotonergic function: a critical review. Life Sci 1993;53: Bancroft J, Cook A, Davidson D, Bennie J, Goodwin G. Blunting of neuroendocrine responses to infusion of L-tryptophan in women with perimenstrual mood changes. Psychol Med 1991;21: Diagnostic and statistical manual of mental disorders. 4th ed. Washington, D.C.: American Psychiatric Association Press, Brammer G. Duodenum is not a consistent source of melatonin in rats. Life Sci 1994;55: Miles LEM, Lipschitz DA, Bleber CP, Cook JD. Measurement of serum ferritin by two-site immunoradiometric assay. Anal Biochem 1974;61: Siegel S. Nonparametric statistics for the behavioral sciences. New York: McGraw Hill, Rabin DS, Schmidt PJ, Campbell G. Hypothalamic-pituitary-adrenal function in patients with the premenstrual syndrome. J Clin Endocrinol Metab 1990;71: Oreland L, Wiberg A, Asberg M. Platelet MAO activity and monoamine metabolites in cerebrospinal fluid in depressed and suicidal patients and healthy controls. Psychiatry Res 1981;4: Coryell W, Schlesser MA. Suicide and the dexamethasone suppression test in unipolar depression. Am J Psychiatry 1981;138: Rapkin AJ, Buckman TD, Sutphin MS, Chang LC, Reading AE. Platelet monoamine oxidase B activity in women with premenstrual syndrome. Am J Obstet Gynecol 1988;159: Roy-Byrne RP, Rubinow DR, Gwirtsman H, Hoban MC, Grover GN. Cortisol response to dexamethasone in women with premenstrual syndrome. Neuropsychobiology 1986;16:61 3. FERTILITY & STERILITY 149