Follicle-stimulating hormone measured in unextracted urine: a reliable tool for easy assessment of ovarian capacity

Transcription

1 FERTILITY AND STERILITY VOL. 70, NO. 3, SEPTEMBER 1998 Copyright 1998 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A. Follicle-stimulating hormone measured in unextracted urine: a reliable tool for easy assessment of ovarian capacity G. Jurjen E. Oosterhuis, M.D.,* Cornelis B. Lambalk, M.D., Ph.D., Henri W. B. Michgelsen, M.D.,* Corry H. De Koning, M.D., Istvan Vermes, M.D., Ph.D.,* and Joop Schoemaker, M.D., Ph.D. Medisch Spectrum Twente Hospital Group, Enschede, and Academic Hospital Free University, Amsterdam, the Netherlands Received January 5, 1998; revised and accepted April 6, Reprint requests: I. Vermes, M.D., Ph.D., Medisch Spectrum Twente Hospital Group, Department of Clinical Chemistry, P.O. BOX 50000, 7500 KA Enschede, the Netherlands (FAX: ; i.vermes@pi.net). * Department of Obstetrics and Gynecology and Department of Clinical Chemistry, Medisch Spectrum Twente Hospital Group. Research Institute for Endocrinology, Reproduction and Metabolism, Academic Hospital Free University /98/$19.00 PII S (98) Objective: To determine the presence of FSH in unextracted urine of perimenopausal women using a microparticle enzyme immunoassay kit on an AxSYM random access immunoassay analyzer. Design: Controlled descriptive study. Setting: A large teaching hospital and infertility clinic. Patient(s): Forty perimenopausal women aged years admitted to our clinic for a gynecological operation. Intervention(s): None. Main Outcome Measure(s): Mean serum FSH level and urinary FSH in early-morning urine, in a random void urine sample, and in 24-hour urine on the same day. FSH in urine on the day of excretion and 1 and 4 weeks thereafter, stored under various conditions. FSH in urine before and after extraction. Result(s): The Pearson s correlation coefficient between mean serum FSH levels and urinary FSH in early morning urine was 0.904, in a random void 0.915, and in 24-hour urine Determination of optimal storage conditions revealed that urine was best kept at 4 C without any additive. The correlation between FSH in extracted and unextracted urine was 98.9%. Conclusion(s): In perimenopausal women, FSH can be reliably measured in unextracted urine. The correlation between urinary FSH and a random void urine sample and mean FSH from a serial serum sample is very high. Urine can be stored for 4 weeks at 4 C without loss of FSH immunoreactivity. (Fertil Steril 1998;70: by American Society for Reproductive Medicine.) Key Words: FSH, urine, gonadotropins, ovarian failure, assay In order to gain information about the functional capacity of the human ovary, the basal FSH level is usually measured in serum on the third day of the menstrual cycle. An elevated serum FSH level is associated with imminent ovarian failure (1 3). The success rate of an IVF treatment is reduced in patients with elevated basal FSH levels (4, 5). In fact, basal FSH level is a better predictor for the outcome of IVF than age (6, 7). The FSH secretions measured in urine are interesting for many reasons. In the first place, there is a substantial variation in basal FSH levels between different menstrual cycles in the same woman (8). This may well be caused by a true variation in FSH levels from cycle to cycle but may also be due to the variation in FSH secretion during the day (9). Therefore, a single serum sample is less reliable to get an impression of FSH secretion. A gold standard for FSH level in serum is a serial sampling. This is an expensive and cumbersome procedure. Urinary FSH levels contain integral information on FSH secretion. In the second place, measuring FSH in serum is an invasive procedure, which decreases patient compliance and, in the case of endocrinological studies, decreases volunteer availability; in contrast, sampling of urine is noninvasive and simple. Several attempts have been made to measure FSH levels in urine. In most studies, a double antibody RIA was used to measure FSH (10 17). A Delfia IFMA (18 20) has also been 544

2 used. The problems in these studies include the need to perform extraction on the urine samples or the type of assay used. These difficulties make the procedure complicated and less suitable for routine use. The aim of this study was to evaluate the possibility of measuring FSH in urine that was not pretreated by using an assay that has been validated for serum FSH. We also determined optimal storage conditions of the urine. We describe a simple and noninvasive method of measuring FSH in nonextracted urine, which can be kept at 4 C for at least 1 month. This technique allows large-scale application and easy access for studies that involve the assessment of the functional capacity of the ovary. SUBJECTS AND METHODS Hormone measurements FSH in urine and serum was measured on an AxSYM random access immunoassay analyzer (Abbott Laboratories, Abbott Park, IL) with an MEIA (microparticle enzyme immuno assay) reagent kit. According to the information provided by the manufacturer, crossreactivity with hcg in this assay is 0.009%, cross-reactivity with LH is 0.2%, and crossreactivity with TSH, 1.2%. Interassay variability is % and intraassay variability is 3.65% 7.64%. The lower detection limit for FSH is 0.09 U/L. In order to correct for dilution variability, creatinine was measured in all urine specimens, and the urinary FSH results were normalized for creatinine concentration by dividing the FSH levels by creatinine levels. Patients For studies involving assay characteristics and storage conditions, we used 24-hour urine samples routinely obtained in our clinic for various purposes from patients of different ages and both sexes. For evaluation of the correlation between serum and urinary FSH, urine and serum samples were obtained from 40 women aged between 32 and 55 years who did not use oral contraceptives and who were admitted to our clinic for a hysterectomy because of a benign illness. The patients were otherwise healthy women with normal renal function. On the day of surgery, we took 6 serum samples (each an hour apart) in order to calculate a mean serum FSH concentration. The first serum sample was taken approximately 1 hour after surgery. During the same day, we collected an early morning urine sample, a 24-hour urine sample, and a random void urine sample. The 24- hour urine sample was initially collected approximately 1 hour before surgery, and collection continued for 24 hours. Specimen handling After obtaining blood by venipuncture, serum was separated from the cells by centrifugation. Serum was either used for measuring FSH immediately, or it was frozen at -20 C until hormone determination. Urine was either collected in a plastic container or in a plastic bag (when collected by catheter). Urine was refrigerated until FSH and creatinine levels were determined, which was done either immediately or within 24 hours. Additional urine samples were taken to determine interassay and intra-assay characteristics, optimal storage conditions, recovery, linearity, and the effect of extraction. To determine optimal storage conditions, we stored urine samples at room temperature, at 4 C, at 20 C without any addition, and at 20 C with addition of 250 L of glycerol in 3 ml of urine. FSH and creatinine levels were measured in these samples after 1 week and 4 weeks. Assay characteristics Extraction was done by adjusting the ph of a urine sample with concentrated glacial acetic acid to 4.5. The sample was mixed with acetone, incubated, and dried on air. The pellet was then resuspended in an assay buffer containing phosphate buffered saline mixed with fetal calf serum, centrifuged, and then the FSH level was determined. The effect of extraction was determined by comparing FSH levels in 15 different urine samples before and after extraction. Linearity was determined by resuspending extracted samples from the same individual in 0.5, 1.0, 1.5, 2.0, and 2.5 ml assay buffers. FSH levels were measured after this resuspension. The observed FSH values were compared with the expected values. Recovery was determined by resuspending extracted samples in six different standard calibrators in solutions containing 0, 1, 10, 50, 100, and 150 U of FSH/L (standard calibrators) or 26, 54, 76, 103, and 124 U of FSH/L (Third International Standard for FSH and LH, urinary). The standard calibrators had been referenced against the World Health Organization Second International Reference Preparation for human FSH. The observed FSH values were compared with the expected values. Ethics All patients and volunteers gave written informed consent before entering the study. The study protocol was approved by the ethics committee of the Medisch Spectrum Twente Hospital Group, Enschede, the Netherlands. Statistics Values are expressed as the means SD. Linear regression analysis was used to analyze the relation between urine and serum FSH concentrations, and Pearson s correlation coefficient was calculated. A P value of.05 was considered statistically significant. FERTILITY & STERILITY 545

3 FIGURE 1 Analytical recovery. FSH concentration is measured in two different urine samples. Both are extracted and diluted in different volumes of assay buffer. FIGURE 3 Correlation between FSH levels in 21 extracted and unextracted urine samples. The correlation coefficient.989 (P.0001). RESULTS Assay characteristics Serial dilutions of urine samples containing high concentrations of FSH showed linear dilution curves (Fig. 1). Correlation between expected and measured FSH levels was 99.95%. The analytical recovery also showed a linear curve (Fig. 2), of the correlation between expected and measured FSH levels was 99.94% for the standard calibrators and 99.03% for the Third International Standard for FSH and LH (urinary). Concordance between FSH concentrations in extracted and unextracted urine samples was high (Fig. 3). The correlation between extracted and unextracted samples was 98.9% (n 21). We therefore decided to use unextracted urine samples subsequently. Storage conditions Decline in FSH immunoreactivity after 1 week was limited, irrespective of storage conditions (Fig. 4). When the sample was stored at 20 C without any additive, 90.9% 26.9 immunoreactivity remained detectable (n 58); when stored at 20 C with glycerol, 86.3% 15.1 immunoreactivity remained detectable (n 58); when stored at 4 C, 98.8% 20.9 immunoreactivity remained detectable (n 18); and when stored at room temperature, 97.2% 20.2 immunoreactivity remained detectable (n 18). FIGURE 2 Expected versus measured FSH levels in urine samples show linear dilution curves. The samples are extracted and then diluted in an assay buffer. Known reference preparations of FSH in different concentrations are added to the solution and FSH level is measured. FIGURE 4 Relative decline in FSH immunoreactivity ( SD) in samples stored at 20 C without additive (x), at 20 C with 250 Lof glycerol in 3 ml of urine (Œ), at 4 C ( ), and at room temperature ( ) after 1 week and 4 weeks. 546 Oosterhuis et al. FSH measured in urine Vol. 70, No. 3, September 1998

4 FIGURE 5 Correlation between mean serum FSH levels and FSH levels in random void urine samples on the same day in 40 perimenopausal women. Pearson s correlation coefficient.915, P After 4 weeks of storage, however, there was a much higher loss of immunoreactivity when the sample was stored at 20 C: without additive only 73.3% 26.7 (n 60) and with glycerol only 80.7% 11.8 (n 60) immunoreactivity remained detectable. Surprisingly, storage at 4 C caused no decline in immunoreactivity after 4 weeks: we found 98.9% 7.7 immunoreactivity measured on the first day (n 20). At room temperature, we found 92.0% 13.5 immunoreactivity compared with that measured on the first day (n 20). Serum versus urinary FSH The correlation between mean serum FSH levels and urinary FSH levels was high. We found a Pearson s correlation coefficient of.915 between mean serum FSH levels and FSH levels in a random void urine sample (n 40, P.0001), of.904 between mean serum FSH levels and FSH levels in an early morning urine sample (n 37, P.0001) and of.857 between mean serum FSH levels and FSH levels in 24-hour samples of urine (n 40, P.0001) (Table 1). DISCUSSION To our knowledge, this is the first report of an assay measuring FSH in unextracted urine, correlating different TABLE 1 Correlation characteristics between mean serum FSH levels and urine FSH levels. Intercept Slope Pearson s correlation coefficient Random void urine sample Early morning urine sample hour urine sample P urine samples with serial serum sampling on the same day, and determining optimal storage conditions as well. Urine samples provide a practical method for monitoring the endocrine function of both patients and volunteers in research programs (21). Urine samples have the advantage of being simple, noninvasive, cost efficient, and stable (22). The high correlation we found between urinary FSH and serial serum FSH levels shows that the assay used is suitable for measuring FSH in urine in a reliable way. So far, there have been no reports about higher correlations between urinary and serum FSH, either when measured by timeresolved Delfia (18, 19) or by double antibody RIA (10, 11, 13, 17, 23). The only report of a correlation between urinary and serum FSH exceeding 90% is that of Chipman et al. (13); these investigators studied a subgroup of normal males, male transsexuals, patients with Turner syndrome, hypogonadism, and amenorrhea from whom serum and urine samples were taken during and after a luteinizing hormonereleasing hormone infusion. Marcus et al. (17) published the single study about determining FSH in urine for the purpose of detecting elevated FSH levels in women with imminent ovarian failure. They found a correlation between urinary FSH levels and serum FSH levels of 70%. We found a very good correlation between serum FSH levels and FSH levels in a random void urine sample, which in routine practice is probably also the easiest to obtain. Other studies have used different types of urine samples to correlate with serum FSH, such as early morning urine samples (16, 19, 20), 3 24 hour samples (10, 11, 13, 14, 23, 24), or random void urine samples (17, 18), but no other study has compared first morning urine, spot urine, and 24-hour urine samples to determine which sample correlates best. We have found no need to use glycerol as a preservative when urine was stored at 20 C before hormone measurement, as did others (25). When FSH levels are determined within 1 week, urine can be stored either at 20 C, at 4 C, or at room temperature without a significant decrease in immunoreactivity. After 4 weeks, the decrease in immunoreactivity in frozen urine is high. Storage at room temperature or at 4 C is most reliable when urine is stored for more than 1 week. For routine practice, it is best to measure FSH in urine within 24 hours. If that is not possible, urine should be stored at 4 C and the FSH level should be determined within 1 month. Because of the good correlation between extracted and unextracted urine samples, we decided to use unextracted samples. This way, the assay is easy to perform on a routine basis in a clinical laboratory setting. The advantage of the assay we describe is low intra-assay and interassay variability, excellent recovery and linearity characteristics, good storage conditions, and high correlations with serum FSH levels. Unlike RIAs or even Delfia assays, this assay is not FERTILITY & STERILITY 547

5 elaborate because there is no need for extraction of the urine samples. Volunteer availability and patient compliance will be improved by the availability of a noninvasive technique for determining ovarian reserve. We believe that the assay we used may improve the sensitivity of infertility diagnosis and reduce patient discomfort. Acknowledgement: The authors would like to thank Abbott Diagnostics, Hoofddorp, the Netherlands, for generously providing all FSH assays. References 1. Buckler HM, Evans CA, Mamtora H, Burger HG, Anderson DC. Gonadotropin, steroid, and inhibin levels in women with incipient ovarian failure during anovulatory and ovulatory rebound cycles. J Clin Endocrinol Metab 1991;72: Ahmed Ebbiary NA, Lenton EA, Salt C, Ward AM, Cooke ID. The significance of elevated basal follicle stimulating hormone in regularly menstruating infertile women. Hum Reprod 1994;9: Cameron IT, O Shea FC, Rolland JM, Hughes EG, De Kretser DM, Healy DL. Occult ovarian failure: A syndrome of infertility, regular menses, and elevated follicle-stimulating hormone concentrations. J Clin Endocrinol Metab 1988;67: Muasher SJ, Oehninger SC, Simonetti S, Matta J, Ellis LM, Liu H-C, et al. The value of basal and/or stimulated serum gonadotropin levels in prediction of stimulation response and in vitro fertilization outcome. Fertil Steril 1988;50: Scott RT, Toner JP, Muasher SJ, Oehninger SC, Robinson S, Rosenwaks Z. Follicle-stimulating hormone levels on cycle day 3 are predictive of in vitro fertilization outcome. Fertil Steril 1989;51: Cahill DJ, Prosser CJ, Wardle PG, Ford WC, Hull MGR. Relative influence of serum follicle stimulating hormone, age and other factors on ovarian response to gonadotrophin stimulation. Br J Obstet Gynaecol 1994;101: Toner JP, Philput CB, Jones GS, Muasher SJ. Basal follicle-stimulating hormone level is a better predictor of in vitro fertilization performance than age. Fertil Steril 1991;55: Van Os HC, Jansen CAM. Intercycle variation of basal FSH and improvement of IVF resluts. Hum Reprod 1992; Filicori M, Butler JP, Crowley WF Jr. Neuroendocrine regulation of the corpus luteum in the human. Evidence for pulsatile progesterone secretion. J Clin Invest 1984;73: Hansen JW, Ross GT. A new method simplifying collection of serial specimens for gonadotropin determinations. J Clin Endocrinol Metab 1975;41: Kulin HE, Bell PM, Santen RJ, Ferber AJ. Integration of pulsatile gonadotropin secretion by timed urinary measurements: an accurate and sensitive 3-hour test. J Clin Endocrinol Metab 1975;40: Ellis MJ, Donald RA, Livesey JH. Simultaneous radioimmunoassay of luteinizing hormone and follicle-stimulating hormone. J Endocrinol 1978;79: Chipman JJ, Moore RJ, Marks JF, Fevre M, Segel T, Ramsey J, et al. Interrelationship of plasma and urinary gonadotropins: correlations for 24 hours, for sleep/wake periods, and for 3 hours after luteinizing hormone-releasing hormone stimulation. J Clin Endocrinol Metab 1981;52: Santner SJ, Santen RJ, Kulin HE, Demers LM. A model for validation of radioimmunoassay kit reagents: measurement of follitropin and lutropin in blood and urine. Clin Chem 1981;27: Landy H, Schneyer AL, Whitcomb RW, Crowley WF. Validation of highly specific and sensitive radioimmunoassays for lutropin, follitropin, and free alpha subunit in unextracted urine. Clin Chem 1990; 36: Maesaka H, Suwa S, Tachibana K, Kikuchi N. Monthly urinary LH and FSH secretory patterns in normal children and patients with sexual disorders. Pediatr Res 1990;28: Marcus M, Grunfeld L, Berkowitz G, Kaplan P, Godbold J. Urinary follicle-stimulating hormone as a biological marker of ovarian toxicity. Fertil Steril 1993;59: Demir A, Alfthan H, Stenman UH, Voutilainen R. A clinically useful method for detecting gonadotropins in children: assessment of luteinizing hormone and follicle- stimulating hormone from urine as an alternative to serum by ultrasensitive time-resolved immunofluorometric assays. Pediatr Res 1994;36: Saketos M, Sharma N, Adel T, Raghuwanshi M, Santoro N. Timeresolved immunofluorometric assay and specimen storage conditions for measuring urinary gonadotropins. Clin Chem 1994;40: Kesner JS, Knecht EA, Krieg EF Jr. Stability of urinary female reproductive hormones stored under various conditions. Reprod Toxicol 1995;9: Lasley BL, Mobed K, Gold EB. The use of urinary hormonal assessments in human studies. Ann N Y Acad Sci 1994;709: Mulder C, Schutter EJM, van Uxem WIM, van Kamp GJ. Stability of human urinary gonadotropin peptide. Tumor Biol 1997;18: Beitins IZ, O Loughlin K, Ostrea T, McArthur JW. Gonadotropin determinations in timed 3-hour urine collections during the menstrual cycle and LHRH testing. J Clin Endocrinol Metab 1976;43: Maesaka H, Suwa S, Tachiba K, Kikuchi N. Quantitation of urinary gonadotropins in normal children. Pediatr Res 1990;28: Livesey JH, Roud HK, Metcalf MG, Donald RA. Glycerol prevents loss of immunoreactive follicle-stimulating hormone and luteinizing hormone from frozen urine. J Endocrinol 1983;98: Oosterhuis et al. FSH measured in urine Vol. 70, No. 3, September 1998