Reports dating from as early as the 1930s have

Transcription

1 Journal of Andrology, Vol. 31, No. 2, March/April 2010 Copyright E American Society of Andrology Assessment of Seminal Estradiol and Testosterone Levels as Predictors of Human Spermatogenesis QIUFANG ZHANG, QUAN BAI, YANG YUAN, PING LIU, AND JIE QIAO From the Center of Reproductive Medicine, Department of Obstetrics and Gynecology, Peking University Third Hospital, Beijing, China. ABSTRACT: The proposed hypothesis for this study was that seminal testosterone/estradiol levels and/or their ratios may be a good indicator for predicting normal spermatogenesis. The concentrations of estradiol and testosterone in seminal fluid were measured using competitive immunoassay techniques in specimens collected from 192 infertile patients and 103 normospermic men. Infertile patients were subdivided into three groups according to their semen analysis results and testicular biopsy: oligozoospermia, obstructive azoospermia (OA), and nonobstructive azoospermia (NOA). Results showed that seminal testosterone levels in the infertile groups were lower than in the normospermic individuals (P,.01), whereas seminal estradiol levels in the OA group were significantly higher than those in normospermic and NOA groups (P,.01). Testosterone/estradiol ratios in the seminal plasma from the infertile groups were significantly lower than that in the normospermic group (P,.01). However, seminal estradiol levels among normospermic and NOA groups showed no significant differences. These results suggest that the local balance between androgen and estrogen, or their ratios, may play an important role in maintaining normal spermatogenesis. Also, decreased seminal testosterone/estradiol ratio may be a good indicator for identifying the absence of sperm production in NOA patients. Key words: Semen, infertility. J Androl 2010;31: Reports dating from as early as the 1930s have indicated that the mammalian testis could synthesize estradiol (Zondek, 1934; Kelch et al, 1972; O Donnell et al, 2001) A later investigation from the early 1970s demonstrated that the human testis does indeed secrete estradiol within the spermatic vein, at a mean concentration that is 50 times greater than that in the peripheral plasma (Kelch et al, 1972). Furthermore, supporting evidence identified estradiol as originating not only from testicular somatic cells, but also from germ cells (Aquila et al, 2002; Carreau et al, 2003). Importantly, investigations upon the rat testis found aromatase mrna and its activity to be higher in the germ cells than in Leydig cells, suggesting that the germ cells are an important source of estradiol (Janulis et al, 1996). These findings led to the hypothesis that local estrogen levels might be associated with the current state of spermatogenesis, or the total number of spermatogenic cells in the testis. In more recent decades, several studies have reported obvious declines in human sperm counts and overall Supported by the national basic research program of China (973 Program, 2007ZB948102). Correspondence to: Dr Jie Qiao, Peking University Third Hospital, No. 49, North Garden Road, Haidian District, Beijing, China ( jie.qiao@263.net). Received for publication January 14, 2009; accepted for publication August 14, DOI: /jandrol quality, accompanied by an increase in male reproductive disorders. Exposure to environmental xenoestradiol has been suggested as one possible cause for these observations (Stillman, 1982; Sharpe and Skakkebaek, 1993; Rozati et al, 2002). Animal experimentation has also demonstrated these environmental estrogenic trends. Rats descended from a great-grandfather who was exposed to high levels of endocrine disruptors during fetal development positively correlated with low levels of sperm production (Anway et al, 2005; Anway and Skinner, 2006; Delbes et al, 2006). Based on these findings, some researchers have proposed a strong connection between local estradiol elevation and spermatogenic failure. Other investigations have revealed an obvious elevation of seminal estradiol in infertile men, particularly in oligozoospermic and oligoteratoasthenozoospermic patients (Bujan et al, 1993; Luboshitzky et al, 2002). However, a variety of causes have been suggested to contribute to oligozoospermia and azoospermia in clinics; these include tract obstruction, inflammation, and spermatogenic failure, to name a few. More recent research has been careful to take into account the fact that most previous studies did not subclassify the patient s etiology into oligozoospermia or oligoteratoasthenozoospermia subtypes (Bujan et al, 1993; Luboshitzky et al, 2002). Strictly speaking, these studies were not critical enough to establish a relationship between the estradiol levels in seminal specimens and 215

2 216 Journal of Andrology N March ÙApril 2010 spermatogenesis. Given the circumstances, certain oligozoospermia cases caused by inflammation or obstruction are not equivalent to complete spermatogenic failure. Apart from estradiol, androgenic influence on sperm production has also spawned much research (Voglmayr et al, 1980). Unfortunately, the relationship between intratesticular androgens and spermatogenesis remained unclear until very recently (Jarow and Zirkin, 2005). It was demonstrated that rat intratesticular testosterone could be reduced by 50% 60% without an adverse effect on spermatogenesis; however, there remained little known regarding the amount of testicular testosterone necessary to maintain normal spermatogenesis in the human. Prior to this investigation, the knowledge of human spermatogenesis in relation to testosterone levels was very narrow. It was known that testosterone could be converted into estrogen within the lumen of the male reproductive tract (Hess, 2000), and that human Leydig cells could express androgen receptors, estrogen receptors, and aromatase. This information suggested that the local balance between estrogenic and androgenic actions may be important for maintaining spermatogenesis (O Donnell et al, 2001). This study attempts to investigate the relationship between local estradiol/testosterone levels, their balance, and the spermatogenesis state. This was accomplished by separating specimens with a more strict grouping method. If the hypothesis is confirmed, these findings may provide a valuable biological indicator for predicting the spermatogenic state. Materials and Methods Population This study was approved by the Ethics Committee of Peking University Third Hospital, China. Written informed consents were obtained from each participant before the initiation of the investigation. 295 male patients between the ages of 24 and 47 years old were prospectively selected from our infertility clinics. Eligibility criteria included that all participants were in good health, were not receiving hormonal treatments, and showed no signs of genital infection. All azoospermic men were followed up with results from their testicular biopsy. Semen Analysis Semen samples were collected by masturbation after 2 5 days of sexual abstinence. Samples were allowed to liquefy at room temperature for minutes and evaluated for sperm count and motility according to World Health Organization (1999) criteria. Following a semen analysis, samples were centrifuged at g for 10 minutes to sediment the spermatozoa. The resulting seminal plasma was immediately separated and divided into 2 aliquots, then stored at 220uC until assayed. According to the results of the semen analysis, patients were divided into three groups: normospermic, consisting of 103 healthy normospermic men, with sperm concentrations $ /ml and normal sperm motility (a $ 25%, ora+ b $ 25%); oligozoospermic, consisting of 94 patients with severe oligozoospermia, with sperm concentrations, /ml; and azoospermic, containing 98 patients with azoospermia, for whom no sperm were found in the ejaculate following centrifugation at g for 10 minutes. Hormone Measurements Seminal plasma samples were thawed at room temperature, sufficiently mixed, then separated into 200-ml aliquots to determine the concentrations of testosterone and estradiol by a competitive immunoassay, employing the Immulite analyzer (Diagnostic Products Corporation, Los Angeles, California). The intra-assay and interassay coefficients of variation were 5.5% and 6.8% for estradiol, and 4.8% and 5.1% for testosterone, respectively. The sensitivities of the assays were 73.4 pmol/l for estradiol and 0.7 nmol/l for testosterone. Following Up in All Azoospermic Patients All azoospermic patients were subdivided into obstructive azoospermia (OA) and nonobstructive azoospermia (NOA) groups. The OA group consisted of patients with a clinical diagnosis of obstruction. Among these, 12 were diagnosed with congenital absence of the vas deferens and seminal vesicles (CAVD). For CAVD patients, semen analysis showed azoospermia with low seminal volume of ml and low ph of 6.5. Semen fructose tests were negative. Transrectal ultrasonography demonstrated that both the seminal vesicles and the vas deferens were critically absent. As such, the OA group was further subdivided into 2 smaller groups: CAVD and non-cavd groups. Patients with incomplete spermatogenesis comprised the NOA group. Histological evidence showed maturation arrest (MA), Sertoli cell only syndrome (SCOS), and tubular sclerosis/atrophy in these patients (Matsumiya et al, 1994). For NOA patients, at least 4 sites were randomly biopsied from each testis. Based on the biopsy results, the NOA group was further subdivided into 2 groups: NOA with sperm (+; NOA with successful sperm recovery, classified by the successful identification of at least 1 spermatozoon in the biopsy); and NOA without sperm (2; NOA with failed sperm recovery, in which no spermatozoon was found in the biopsy from either side of the testis). Statistical Analysis For this study, all statistical analyses were carried out using SPSS 10.0 for Windows; results were expressed as a mean 6 SD. Statistically significant differences between mean values were evaluated using a 1-way analysis of variance; a post hoc test (LSD or Tamhane s T2) was used for multiple comparisons. Statistically significant differences between subgroups were compared using independent sample tests. Correlations between study parameters were made by nonparametric

3 Zhang et al N Seminal Estradiol and Testosterone in Fertile Men 217 Table 1. Seminal plasma hormone concentrations and their ratios in normospermic, oligozoospermic, and azoospermic groups Norm Oligo Az N Age, y Estradiol, pmol/l a Testosterone, nmol/l b c T/E 2 ratio d e Abbreviations: Az, azoospermic group; Norm, normospermic group; Oligo, oligozoospermic group; T/E 2, testosterone/estradiol. a P 5.008, Az vs Norm. b P 5.005, Oligo vs Az. c P 5.000, Az vs Norm. d P 5.003, Oligo vs Norm. e P 5.000, Az vs Norm; P 5.01, Az vs Oligo. Spearman s correlation coefficient. A P value of,.05 was considered statistically significant for all analyses. Receiver operating characteristic (ROC) curves were constructed to examine the predictive value of seminal hormone levels and their ratios. Sensitivity (y-axis) against (1 2 specificity) (x-axis) was plotted at each threshold level and the area under the curve (AUC) was computed by nonparametric Wilcoxon testing. The AUC value represents the predictive value for successful sperm recovery before a testicular biopsy among NOA patients. A value of,0.5 represents no increased chance of recovering spermatozoa following surgery. Results Seminal Plasma Hormone Levels The results (mean 6 SD) of seminal plasma testosterone, estradiol, and testosterone/estradiol (T/E 2 ) ratios are given in Table 1. Patient age was not significantly different among the three groups. Seminal plasma estradiol levels in the azoospermic group were significantly higher than levels in normospermic men (P,.01); however, there were no significant differences in estradiol levels between either azoospermic and oligozoospermic groups, or oligozoospermic and normospermic groups. Seminal plasma testosterone concentrations were significantly lower in the azoospermic group as compared with normospermic (P,.01) and oligozoospermic groups (P,.01). All three groups demonstrated significant differences in seminal plasma T/E 2 ratio (P,.05). Seminal Plasma Hormone Levels in Azoospermic Patients After following up all azoospermic diagnoses, a total of 79 patients (23 OA and 56 NOA) had either a testicular biopsy or epididymal aspiration performed. In the NOA group, there were 13 patients with hypospermatogenesis, whereas the other 43 men were diagnosed with MA or with SCOS (absence of sperm). The NOA group had lower seminal testosterone concentrations than both the normospermic group and the OA group (P,.01; Table 2); all NOA patients with either MA or SCOS also had lower testosterone levels and T/E 2 ratios (P,.001). In the OA group, there were 12 patients with CAVD, and the other 11 men were without CAVD. The OA group demonstrated significantly higher estradiol levels and lower T/E 2 ratios (P,.001) as compared with the normospermic and NOA groups. In addition, the CAVD group had much higher estradiol levels ( pmol/l) and lower T/E 2 ratio values ( ) than other OA and NOA patients. After CAVD patients were excluded from the OA group, seminal estradiol levels were still significantly higher in the OA group vs the NOA group ( vs , P,.001; Table 2; Figure 1). Predictive Value of Estradiol, Testosterone, and Their Ratios in Biopsy-Based Diagnoses For all NOA patients, there were significant positive correlations between biopsy diagnoses (presence or absence of sperm) and testosterone levels within the seminal plasma (Spearman s r , P 5.006), as well as T/E 2 ratios (Spearman s r , P 5.000). Conversely, there were no significant correlations between biopsy diagnoses and estradiol levels (Spearman s r , P 5.719) (Table 3). The predictive potency of seminal estradiol, testosterone, and their ratios was tested by ROC mapping. As displayed in Figure 2, the area under the ROC curve for testosterone peaked at a value of (95% confidence interval [CI], ), and the T/E 2 ratio peaked at (95% CI, ). Several threshold values of seminal testosterone levels and T/E 2 ratios were analyzed in terms of their specificity and sensitivity. Table 4 visualizes that the best compromise between specificity (81.8%, 84.8%) and sensitivity (72.7%, 72.7%) can be obtained with cutoff values of 4.10

4 218 Journal of Andrology N March ÙApril 2010 Table 2. Seminal plasma hormone concentrations and their ratio in normospermic, obstructive azoospermic, and nonobstructive azoospermic groups (data are mean 6 SD) Norm OA NOA N Estradiol, pmol/l a Testosterone, nmol/l b T/E 2 ratio c d Abbreviations: NOA, nonobstructive azoospermic group; Norm, normospermic group; OA, obstructive azoospermic group; T/E 2, testosterone/ estradiol. a P 5.000, OA vs Norm or NOA. b P 5.000, NOA vs Norm. c P 5.000, OA vs Norm; P 5.015, OA vs NOA. d P 5.001, NOA vs Norm group. nmol/l for seminal testosterone and for seminal T/E 2 ratio, respectively. According to the established cutoff values for seminal testosterone and seminal T/E 2 ratio (4.1 and , respectively), 29 of 56 NOA patients (51.8%) were predicted to fail surgical sperm recovery; of these 29 predictions, 3 patients were found with spermatozoon in their testes. This data establishes the rate of missed diagnosis to be 10.3% (3 of 29). Additionally, of the 13 of 56 NOA patients (23.21%) predicted to have successful surgical sperm recovery, 3 were found without spermatozoa in their testes, setting the misdiagnosis rate at 23.08%. And for the other 14 patients, because of the mismatched values (one was less than the cutoff value, and another was more than the cutoff value), we could not predict their testicular biopsy. Finally, we found 7 men with spermatozoa and 7 men without spermatozoa in their testes. Discussion Figure 1. Seminal plasma hormone concentrations and their ratios in azoospermics (data are mean 6 SD). *, P 5.000: CAVD group vs OA without CAVD group or NOA group; n, P,.01: NOA with sperm (2) group vs OA without CAVD group or NOA with sperm (+) group; q, P 5.000: NOA with sperm (2) group vs NOA with sperm (+) group; ww, P 5.000: CAVD group vs OA without CAVD group or NOA group; CAVD, OA patients with congenital absence of the vas deferens and seminal vesicles; OA, obstructive azoospermia; NOA, nonobstructive azoospermia; NOA with sperm +, NOA patients with successful sperm recovery group; NOA with sperm 2, NOA patients with failed sperm recovery group. For those men included in this study, the data shown in Table 1 demonstrate significantly higher seminal plasma estradiol levels as compared with normospermic men, who were also representative of previous reports (Bujan et al, 1993; Luboshitzky et al, 2002). The current data also suggest that azoospermic patients have extraordinarily high seminal estradiol levels, which can potentially diagnose CAVD. There are 2 possible reasons for the elevated seminal estradiol levels in these patients: 1) estrogen receptor expression in the whole testis is dramatically decreased in MA and SCOS patients (Han et al, 2009), which may reduce estrogen receptor binding and indirectly increase seminal estradiol levels; 2) seminal estradiol may concentrate in the prostate, resulting in an increased seminal estradiol level. This finding is supported by previous work by Risbridger et al (2003).

5 Zhang et al N Seminal Estradiol and Testosterone in Fertile Men 219 Table 3. Correlation coefficients relating sperm retrieval to seminal plasma hormone concentrations and their ratio in failed-biopsy and successful-biopsy groups R P Estradiol, pmol/l NS Testosterone, nmol/l Testosterone/estradiol ratio Estrogen and androgen can be considered 2 sides of the same coin. On one hand, androgens can be converted into estrogens; on the other, Sertoli cells are known to secrete fluid under androgen control, and estrogen is involved in the resorption of the fluid in the efferent ducts (Hess et al, 1997; Sharpe, 1997). Despite a lack of knowledge about the specific mechanisms of how estrogen, androgen, or their balance affects spermatogenesis, this data clearly indicates that their local balance plays an important role in maintaining physiological spermatogenesis. Our results also suggest that seminal T/E 2 ratios in OA patients, especially those diagnosed with CAVD, may be a good indicator for predicting azoospermia. Although some studies have claimed that serum testosterone levels demonstrate no relationship to sperm concentration or testicular biopsy (Jackaman et al, 1977; Table 4. ROC curve data Threshold Level Sensitivity, % Specificity, % Testosterone, nmol/l Testosterone/estradiol ratio Goulis et al, 2008), others suggest that testosterone may assume a critical role in both morphological development and reproductive function in the male (Holdcraft and Braun, 2004; Takada et al, 2008). Treatment with exogenous testosterone has been shown to fully rescue spermatogenesis in LH-R knockout mice (Lei et al, 2001). In this study, NOA patients without sperm (2) show very low levels of seminal testosterone, as compared with NOA patients with sperm (+); their biopsy tissues indicate severe seminiferous tubule atrophy, sclerosis, and Leydig cell hyperplasia. These findings imply that seminal testosterone levels may predict whether normal spermatogenesis is occurring within the seminiferous tubules of NOA patients. In addition, serum hormone levels (prolactin [PRL], luteinizing hormone [LH], follicle-stimulating hormone [FSH], estradiol, and testosterone) were analyzed from NOA patients; the results suggest that FSH levels are significantly different between groups with sperm (+) and without sperm (2) ( IU/L vs IU/L, respectively). However, testosterone levels were not significantly deferent between the 2 groups ( nmol/l vs nmol/l, respectively; unpublished data). Therefore, FSH levels in serum may be considered as another indicator to predict normal spermatogenesis in NOA patients (Goulis et al, 2008). In summary, these results demonstrated that 1) reduced T/E 2 ratios and increased seminal estradiol levels may reflect the type of obstruction or CAVD, and 2) T/E 2 ratios coupled with seminal testosterone levels may serve as good indicators for predicting the success of surgically retrieving spermatozoa from the testes from nonobstructive azoospermic patients. However, additional research encompassing a larger patient population is needed to confirm these predictive values in helping to avoid unnecessary testicular biopsies. Figure 2. ROC curve for seminal plasma hormone concentrations and their ratio for prediction of the presence of spermatozoon from testis of NOA patients. Acknowledgments We appreciate Chisun Yang for excellent technical assistance, Qian Chen and Renpei Yuan for sample collecting, and Jing Qiao Lv for kind assistance in data analyses. Appreciation is extended to Dr Huai L. Feng and Mr Dennis E. Marchesi for their critical review and comments.

6 220 Journal of Andrology N March ÙApril 2010 References Anway MD, Cupp AS, Uzumcu M, Skinner MK. Epigenetic transgenerational actions of endocrine disruptors and male fertility. Science. 2005;308(5727): Anway MD, Skinner MK. Epigenetic transgenerational actions of endocrine disruptors. Endocrinology. 2006;147(6 Suppl):S43 S49. Aquila S, Sisci D, Gentile M, Middea E, Siciliano L, Ando S. Human ejaculated spermatozoa contain active P450 aromatase. J Clin Endocrinol Metab. 2002;87(7): Bujan L, Mieusset R, Audran F, Lumbroso S, Sultan C. Increased oestradiol level in seminal plasma in infertile men. Hum Reprod. 1993;8(1): Carreau S, Lambard S, Delalande C, Denis-Galeraud I, Bilinska B, Bourguiba S. Aromatase expression and role of estrogens in male gonad: a review. Reprod Biol Endocrinol. 2003;1:35. Delbes G, Levacher C, Habert R. Estrogen effects on fetal and neonatal testicular development. Reproduction. 2006;132(4): Goulis DG, Tsametis C, Iliadou PK, Polychronou P, Kantartzi PD, Tarlatzis BC, Bontis IN, Papadimas I. Serum inhibin B and antimullerian hormone are not superior to follicle-stimulating hormone as predictors of the presence of sperm in testicular fineneedle aspiration in men with azoospermia. Fertil Steril. 2009;91: Han Y, Feng HL, Sandlow JI, Haines CJ. Comparing expression of progesterone and estrogen receptors in testicular tissue from men with obstructive and nonobstructive azoospermia. J Androl. 2009;30(2): Hess RA. Oestrogen in fluid transport in efferent ducts of the male reproductive tract. Rev Reprod. 2000;5(2): Hess RA, Bunick D, Lee KH, Bahr J, Taylor JA, Korach KS, Lubahn DB. A role for oestrogens in the male reproductive system. Nature. 1997;390: Holdcraft RW, Braun RE. Hormonal regulation of spermatogenesis. Int J Androl. 2004;27(6): Jackaman R, Ghanadian R, Ansell ID, McLoughlin PV, Chisholm GD. Relationships between spermatogenesis and serum hormone levels in subfertile men. Br J Obstet Gynaecol. 1977;84(9): Janulis L, Bahr JM, Hess RA, Bunick D. P450 aromatase messenger ribonucleic acid expression in male rat germ cells: detection by reverse transcription-polymerase chain reaction amplification. J Androl. 1996;17(6): Jarow JP, Zirkin BR. The androgen microenvironment of the human testis and hormonal control of spermatogenesis. Ann N Y Acad Sci. 2005;1061: Kelch RP, Jenner MR, Weinstein R, Kaplan SL, Grumbach MM. Estradiol and testosterone secretion by human, simian, and canine testes, in males with hypogonadism and in male pseudohermaphrodites with the feminizing testes syndrome. J Clin Invest. 1972;51(4): Lei ZM, Mishra S, Zou W, Xu B, Foltz M, Li X, Rao CV. Targeted disruption of luteinizing hormone/human chorionic gonadotropin receptor gene. Mol Endocrinol. 2001;15(1): Luboshitzky R, Kaplan-Zverling M, Shen-Orr Z, Nave R, Herer P. Seminal plasma androgen/oestrogen balance in infertile men. Int J Androl. 2002;25(6): Matsumiya K, Namiki M, Takahara S, Kondoh N, Takada S, Kiyohara H, Okuyama A. Clinical study of azoospermia. Int J Androl. 1994;17(3): O Donnell L, Robertson KM, Jones ME, Simpson ER. Estrogen and spermatogenesis. Endocr Rev. 2001;22(3): Risbridger GP, Bianco JJ, Ellem SJ, McPherson SJ. Oestrogens and prostate cancer. Endocr Relat Cancer. 2003;10(2): Rozati R, Reddy PP, Reddanna P, Mujtaba R. Role of environmental estrogens in the deterioration of male factor fertility. Fertil Steril. 2002;78(6): Sharpe RM. Do males rely on female hormones? Nature. 1997; 390(6659): Sharpe RM, Skakkebaek NE. Are oestrogens involved in falling sperm counts and disorders of the male reproductive tract? Lancet. 1993;341(8857): Stillman RJ. In utero exposure to diethylstilbestrol: adverse effects on the reproductive tract and reproductive performance and male and female offspring. Am J Obstet Gynecol. 1982;142(7): Takada S, Tsujimura A, Ueda T, Matsuoka Y, Takao T, Miyagawa Y, Koga M, Takeyama M, Okamoto Y, Matsumiya K, Fujioka H, Nonomura N, Okuyama A. Androgen decline in patients with nonobstructive azoospermia after microdissection testicular sperm extraction. Urology. 2008;72(1): Voglmayr JK, Roberson C, Musto NA. Comparison of androgen levels in ram rete testis fluid, testicular lymph and spermatic venous blood plasma: evidence for a regulatory mechanism in the seminiferous tubules. Biol Reprod. 1980;23(1): World Health Organization. WHO Laboratory Manual for the Examination of Human Semen and Sperm-Cervical Mucus Interactions. 3rd ed. Cambridge, United Kingdom: Cambridge University Press; 1999:4 33. Zondek B. Mass excretion of estrogenic hormone in the urine of the stallion. Nature. 1934;33: