and Policlinic for Obstetrics and Gynecology of the University, University of Münster, Münster, Germany

Transcription

1 Elevated follicle-stimulating hormone levels and the chances for azoospermic men to become fathers after retrieval of elongated spermatids from cryopreserved testicular tissue Michael Zitzmann, M.D., Ph.D., a Verena Nordhoff, Ph.D., a Victoria von Schönfeld, Ph.D., a Annette Nordsiek-Mengede, M.D., a Sabine Kliesch, M.D., Ph.D., b Andreas N. Schüring, M.D., c Craig Marc Luetjens, Ph.D., a Axel Kamischke, M.D., Ph.D., a Trevor Cooper, Ph.D., a Manuela Simoni, M.D., Ph.D., a and Eberhard Nieschlag, M.D., Ph.D. a a Institute of Reproductive Medicine of the University, b Clinic and Policlinic for Urology of the University, and c Clinic and Policlinic for Obstetrics and Gynecology of the University, University of Münster, Münster, Germany Objective: To assess individual chances for a live-born child in azoospermic men by performance of testicular sperm extraction (TESE) followed by intracytoplasmatic sperm injection (ICSI). Design: A retrospective cohort study. Setting: An academic fertility care center and research unit. Patient(s): Two hundred three couples who wished to have a child; all men had azoospermia. Intervention(s): All men were operated for TESE; 112 men were found to have elongated spermatids (ES), and 209 ICSI cycles were performed in these men using cryopreserved tissue. Main Outcome Measure(s): Predictors for the chances to obtain live sperm and for probabilities of fertilization, clinical pregnancies, and live births. Result(s): Testicular volume, FSH, and inhibin B levels were predictors for the presence of ES. Intracytoplasmic sperm injection resulted in 23 pregnancies, leading to 20 live births. Despite the presence of ES and performance of ICSI in cases of FSH levels 20 IU/L, no pregnancy resulted in these men (n 21). Receiver operating characteristics revealed FSH levels of 20 IU/L as cutoff for treatment success. The number of testicular tubuli containing ES served as a predictor for clinical pregnancy as well as for live birth. Cigarette smoking by the male partner exerted a significant negative influence on treatment success. Conclusion(s): The degree of completely maintained spermatogenesis within the biopsy appears to reflect intrinsic abilities of spermatozoa to induce normal embryo development. Charts based on regression models are presented for counseling patients before TESE; these explain chances of finding ES and probability of successful ICSI. Obtaining offspring is unlikely in cases of azoospermia and of FSH levels of 20 IU/L. (Fertil Steril 2006;86: by American Society for Reproductive Medicine.) Key Words: Intracytoplasmic sperm injection, ICSI, testicular sperm extraction, TESE, assisted reproduction, success rate, predictors, FSH Received August 19, 2005; revised and accepted December 16, Reprint requests: Eberhard Nieschlag, M.D., Ph.D., Institute of Reproductive Medicine of the University, Domagkstrasse 11, D Münster, Germany (FAX: ; eberhard.nieschlag@ ukmuenster.de). The first child conceived by intracytoplasmic sperm injection (ICSI) after testicular sperm extraction (TESE) was born in 1995 (1). This achievement offered patients with azoospermia a reasonable chance for a child of their own (2 6). However, accurate estimation of chances for such a successful procedure remains elusive, as does prediction of the presence of elongated spermatids in testicular tissue by noninvasive parameters, thus making reasonable assessment and counseling before the operation difficult. Several studies relate the outcome of ICSI to testicular sperm from patients with either obstructive or nonobstructive azoospermia. Although some investigators found significantly reduced fertilization rates and also found reduced pregnancy rates in nonobstructive cases (7 9), other investigators described a lower fertilization rate without effect on the pregnancy rate (6, 10, 11) or no significance in fertilization or pregnancy rates at all (12). The definition of azoospermia as obstructive is clearcut in clinical terms: normal testicular volume, normal FSH levels, and reduced seminal markers of the accessory glands. These requirements are met most often in cases of postvasectomy or congenital bilateral aplasia of the vasa deferentes (CBAVD). Nevertheless, not every patient presenting with azoospermia can have his disorder designated definitely as obstructive or nonobstructive, because clinical and histological pictures often exhibit a mixed pattern. Thus, dividing patients into two groups may impede data evaluation. Azoospermic men with deletions of genes in the AZF regions of the Y chromosome may be considered nonobstruc /06/$32.00 Fertility and Sterility Vol. 86, No. 2, August 2006 doi: /j.fertnstert Copyright 2006 American Society for Reproductive Medicine, Published by Elsevier Inc. 339

2 tive, but can, nevertheless, exhibit clinical features such as normal FSH concentrations and normal testis size, thus mimicking an obstructive cause (13). Also, vasectomy, the most obvious cause of obstructive azoospermia, can be accompanied by a significant impairment of spermatogenic efficacy. This time-dependent effect also can cause FSH levels of vasectomized men to increase with the duration of the obstructive interval, even into the supernormal range (14). Rather, parameters describing azoospermia, such as histological findings, as well as noninvasive measures such as testicular volume and FSH levels, constitute a continuous spectrum comprising the diagnosis of azoospermia. We have demonstrated elsewhere that noninvasive parameters are associated with histological findings in testicular tissue (15). In contrast to the use of descriptive terms, we examined how a histologically confirmed degree of spermatogenic impairment (representing a continuous variable) relates to the outcome of ICSI that is performed with TESE samples. The hypothesis of our study was that the outcome of ICSI with spermatids from tissue samples with poor spermatogenesis would be worse than would results from samples with more effective spermatogenesis; this would confirm that impaired fertilization abilities and aberrant embryo development are a function of the quality and degree of spermatogenesis. Endpoints of this investigation were presence of elongated spermatids, fertilization rate of oocytes, and clinical pregnancies and live births. MATERIALS AND METHODS Patients Two hundred three azoospermic men underwent testicular biopsies for histology and TESE, if possible. In 91 men, no elongated spermatids were found. In the remaining 112 couples, 209 TESE-ICSI cycles were performed. Testicular tissue was cryopreserved in all cases, and in the TESE cycles, only elongated spermatids were used. The use of fresh testicular material for ICSI implies the preoperation hormonal stimulation of the female partner, with all its financial burdens and physical risks. Because only about 50% of our patients have elongated spermatids and we cannot completely predict the presence of these cells, we refrain from using fresh biopsy material. We only perform such a procedure in men who have sperm previously present in the ejaculate but have no useful sperm found at the time of ovum pickup. Only then would an emergency TESE be performed and fresh biopsy material be used. Such patients are rare and are not included in this analysis. Earlier, all male patients underwent a clinical examination that included ultrasonography of scrotal contents, hormones, and semen parameters (16). Female partners had provided blood for hormone analysis in previous cycles and underwent a physical examination; all women had apparently normal reproductive functions. All couples were counseled by human geneticists. All patients gave written informed consent for the use of their data for scientific evaluation as approved by the ethics committee (local institutional review board) of the University of Münster and by the State Medical Board, Münster, Germany. Considering the success rates, it has to be kept in mind that in Germany, a maximum of three embryos can be produced, and embryo selection is against the federal embryo protection law. Levels of FSH Serum levels of FSH were determined by highly specific time-resolved fluoroimmunoassays (Autodelfia, Perkin-Elmer, TABLE 1 Diagnostic overview of the 203 azoospermic men involved in this study, data as median and (range; not normally distributed values). Diagnosis No. of elongated spermatids present/total % of Tubuli with complete spermatogenesis in the more intact testis FSH in IU/L Inhibin B in pg/ml Idiopathic azoospermia 95/ (0 100) 11.4 ( ) 81 ( ) AZFc deletion of the Y 0/5 0 (0 0) 15.1 ( ) 50 (22 100) chromosome Klinefelter s syndrome 0/2 0 (0 0) 14.7 ( ) 30 (13 47) CBAVD 7/7 87 (7 98) 3.3 ( ) 142 (49 199) Ejaculation disorders 6/6 42 (29 94) 4.5 ( ) 115 (25 154) Postvasectomy 4/4 88 (66 93) 6.1 ( ) 98 (40 199) All 112/ (0 100) 9.4 ( ) 83 ( ) Note: AZFc ; ejac ejaculate. 340 Zitzmann et al. Chances for fatherhood in azoospermia Vol. 86, No. 2, August 2006

3 Wallac Inc., Turku, Finland). The lower detection limit for FSH was 0.25 IU/L, and imprecision was below 5%. Inhibin B Serum levels were measured by using a commercially available, double-antibody, enzyme-linked immunoassay (Serotec Ltd., Oxford, Oxfordshire, UK). Intraassay and interassay coefficients of variation were 4% and 20%, respectively. The sensitivity of the inhibin B assay, defined as the value of 2 SD above the mean of 10 repeated measurements that were performed with the 0 standard, was 7.8 pg/ml, a concentration corresponding to that of the lowest point of the standard curve. The normal range of serum inhibin B concentrations obtained from values in 84 proven fathers was pg/ml (16). Semen Analysis Semen analysis was performed according to World Health Organization guidelines (17). Semen volume was measured in a graduated glass cylinder. Neutral -glucosidase, zinc, and fructose were measured by multiwell spectrophotometric assays (17). Samples were measured in duplicate, together with quality control pooled samples containing medium concentrations of markers that were included at the beginning and end of the assay to monitor drift. Assay results were accepted when duplicates agreed within 10%, drift was 15%, and internal quality control samples fell within 2 SD of the mean preassay values. Lower limits of reference in our laboratory are 13 mol per ejaculate for fructose, 20 mu per ejaculate for -glucosidase, and 2.4 mol per ejaculate for zinc (17 20). Testicular Volume Sonographic (ultrasound scanner type 2002 ADI; BK Medical, Gentofte, Denmark) measurements of the testes were performed by applying a high-frequency 7.5-MHz convex scanner. Volume was calculated by using the ellipsoid method (16). The volume of the larger testis was considered in analyses involving noninvasive parameters. In cases of statistical analyses using biopsy material, the volume of the respective testis was used. Testicular Biopsy Testicular biopsies were performed after scrotal incision and preparation of the tunica albuginea of the testis. Six biopsies were taken from the same incision of the tunica albuginea. Each biopsy was about 5 mm in length and 2 mm in diameter, taken with scissors, picked up by a needle without touching the biopsy specimen and immediately transferred to the fixatives (Bouin s or Stieve s solution) or medium. Preparation of the Fresh Testicular Biopsy Enzymatic digestion with collagenase at the time of biopsy is a procedure that is used to obtain instant information for the patient; additional information concerning sperm motility is obtained before the ICSI procedure. One testicular biopsy of each side was enzymatically digested with collagenase (0.8 mg/ml, Sigma) for 2 hours and processed by centrifuge at 200 g for 10 minutes. The pellet was dissolved in L of Sperm Preparation Medium (Medicult, Aarhus, Denmark). Of this suspension, 20 L was analyzed on a slide, and each microscopic field was screened for the appearance of spermatozoa at a magnification of 400. The total number of spermatozoa was calculated from the number per microscopic field or was counted per whole slide. Motility was documented as total motility, as a percentage. All other testicular samples were cryopreserved. Histology A second sample from each side was analyzed histologically. The tissue was fixed in Bouin s solution for TABLE 1 Continued. -Glucosidase in (mu/ejac) Zinc ( mol/ejac) Fructose ( mol/ejac) Volume of larger testis 32 (2 219) 6 (0 18) 53 (1 194) 17 (2 60) 42 (20 72) 4 (2 12) 41 (11 116) 15 (10 16) 21 (20 42) 6 (5 8) 40.0 (37 43) 2 (1 3) 7 (3 11) 4 (3 5) 1 (0 66) 23 (17 45) Not obtained Not obtained Not obtained 19 (12 24) 11 (5 21) 3 (2 10) 20 (1 34) 26 (17 28) 32 (2 219) 6 (0 18) 48 (0 194) 17 (2 60) Fertility and Sterility 341

4 hours and was washed in 70% alcohol. Sections of 4 5 m were prepared and stained with PAS, and histological analysis was performed. The number of elongated spermatids in 100 or more individual tubuli was counted in each section from each side. According to our understanding, elongated spermatids represent the highest degree of maturity that spermatozoa can achieve within the testis. Elongated spermatids were used throughout the ICSI procedures that were described and analyzed in this article. In contrast to elongated spermatids, we believe that the term mature spermatozoa reflects the completely developed sperm, which is ejaculated and has undergone further processes of maturation in the epidymidis. Because ejaculated sperm and sperm obtained from epididymal aspiration were not used in the patients described here, we refrain from using the term mature spermatozoa. Nevertheless, we are aware that some publications also apply this term to sperm found in testicular biopsies. The degree of completely maintained spermatogenesis within the biopsy is expressed as a percentage of tubuli containing elongated spermatids in relation to all tubuli that were assessed during histological evaluation. Usually, 100 tubuli per testis were investigated in this way. Testicular Sperm Extraction, ICSI, and Embryo Transfer All female partners were treated with agonist or antagonist protocols (Decapeptyl [Ferring, Germany], Synarela [Pharmacia, Germany], and Cetrotide [Serono, Switzerland]), followed by ovarian stimulation with recombinant FSH (Gonal F, Serono; Puregon, Organon, the Netherlands). Ovulation was induced by a single injection of hcg (Choragon, Ferring), hours before ovum pickup. Oocyte cumulus complexes were retrieved by ultrasound-guided puncture of the follicles and were collected in drops of universal IVF medium (Medicult). Cumulus removal was performed by hyaluronidase treatment (80 IU/mL, Sigma) for 30 seconds and by pipetting up and down in a Pasteur pipette. Oocytes were incubated at 37 C, 5% CO 2, and 95% humidity up to sperm injection. Procedures followed standards published elsewhere (21, 22). Testicular biopsies were thawed and enzymatically digested with collagenase (0.8 mg/ml, Sigma) for 2 hours and processed by centrifuge at 200 g for 10 minutes, and the pellet was suspended in L of sterile medium (Medicult). Two microliters of this suspension was filled into 3- L drops of medium containing 0 3 L of polyvinyl pyrrolidone (PVP) in a sterile mineral oil covered Petri dish. After settling of the cells, each drop was controlled for the occurrence of elongated spermatids by screening each microscopic field at 400 magnification. A single spermatozoon was taken up into an ICSI pipette (Humagen, Charlottesville, SC) and was transferred to a collecting drop. Motile and morphologically normal sperm were preferred, but in case of only immotile or abnormal sperm, these were used for standard ICSI. Maximal collection time did not exceed 4 hours. TABLE 2 Baseline data and comparisons of 203 azoospermic men (between groups positive and negative for elongated spermatids). Parameter All patients Positive for elongated spermatids Negative for elongated spermatids Mann-Whitney test P n Age (y) 35 (20 63) 27 (23 63) 33 (20 54).001 FSH (IU/L) 9.4 ( ) 4.8 ( ) 17 ( ).001 Inhibin B (pg/ml) 83 ( ) 100 ( ) 40 ( ).001 Volume of larger testis (ml) 17 (2 60) 22 (4 60) 13 (2 36).001 -Glucosidase (mu/ejac) 32 (2 219) 24 (2 173) 43 (3 219).002 Zinc ( mol/ejac) 6 (0 18) 5 (1 17) 7 (0 18).04 Fructose ( mol/ejac) 48 (0 194) 40 (0 178) 63 (3 194).002 % of Tubuli with elongated 29 (0 100) 70 (1 100) spermatids No. of motile sperm in the 1,600 (1 20,000) fresh biopsy material Grade A motility of sperm in the fresh biopsy material (%) 10 (1 60) Note: Data are median and range. Ejac ejaculate. 342 Zitzmann et al. Chances for fatherhood in azoospermia Vol. 86, No. 2, August 2006

5 FIGURE 1 Serum FSH concentrations in relation to volume of the larger testis (n 203). Markers give the information whether testicular biopsies contained elongated spermatids (, n 112) or not (, n 91). Note the continuous distribution of values with an absence of clusters. Injection was performed 4 10 hours after oocyte retrieval. In accordance with German embryo protection law, a maximum of three pronuclear stages were chosen for continued culture. The remaining pronuclei had to be frozen or discarded. Embryo transfer was performed hours after oocyte retrieval. Confirmation of Pregnancy Clinical pregnancy cycles were defined as cycles in which a viable pregnancy could be verified by the presence of at least one gestational sac in which a fetal heartbeat could be seen on transvaginal ultrasound examination 6 weeks after embryo transfer. Confirmation of full-term live birth was obtained by standardized telephone interviews of couples registered as clinically pregnant. Statistics Nonparametric Mann-Whitney tests were used for simple comparisons between outcome groups. For the purpose of backwards stepwise binomial regression analysis, all values were log-transformed for analysis because of skewed distribution, except percentage values, which were arc-sin transformed. Repetition of treatment or first treatment cycle were coded as dummy variable. When treatment was repeated, the parameters of the samples of the testis used were applied. Probability curves were generated by using the formula P 1/(1 e z ), with z constant regression coefficient parameter (e.g., FSH level). Degree of fit of the regression models was expressed as Nagelkerke s R 2. Computations were performed by using the SPSS statistical software package (release ; SPSS, Chicago, IL). Unless otherwise stated, results are given as mean SD in tables and figures. Two-sided P values of.05 were considered significant. RESULTS A diagnostic overview of all patients is given in Table 1. Further patient characteristics and baseline parameters with regard to the presence or absence of elongated spermatids are given in Table 2. In addition, the distribution of patients with existing or absent spermatogenesis in relation to serum FSH concentrations and the volume of the larger testis are shown in Figure 1. A backwards stepwise binomial regression model excluded -glucosidase, fructose, and zinc content of the ejaculate and confirmed serum FSH concentrations and volume of the larger testis as significant and independent predictors of the presence of elongated spermatids (both P.001). Nagelkerke s R 2 was For counseling purposes, regression coefficients of FSH and testis volume were used to create probability curves for sperm retrieval (Fig. 2). The association of inhibin B concentrations with FSH levels was r 0.56 (P.001) in men with elongated spermatids and was r 0.74 (P.001) in men with absence of elongated FIGURE 2 Probability curves to predict chances of obtaining elongated spermatids from testicular biopsies using noninvasive parameters. The x axis displays FSH concentrations; the y axis, the probability of finding sperm by TESE. The different curves are derived from various testicular volumes and give the volume in milliliters. Fertility and Sterility 343

6 TABLE 3 Comparisons of noninvasive presurgical parameters in 203 azoospermic men undergoing TESE, partially followed by ICSI (209 cycles in 112 men), concerning the ultimate outcome of fertility treatment. Parameter (per treatment attempt) Yes No Mann-Whitney test (P) Clinical pregnancies n Male partner s age (y) 37, , Female partner s age (y) 33, , FSH (IU/L) 4.5, , Inhibin B (pg/ml) 101, , Volume of larger testis (ml) 20, , Glucosidase (mu/ejac) 29, , Zinc ( mol/ejac) 5, 2 8 6, Fructose ( mol/ejac) 34, , Male smoking, n (%) 4 (17.4) 121 (43.7).014 a Live births n Male partner s age (y) 37, , Female partner s age (y) 34, , FSH (IU/L) 4.4, , Inhibin B (pg/ml) 101, , Volume of larger testis (ml) 20, , Glucosidase (mu/ejac) 29, , Zinc ( mol/ejac) 5, 2 8 6, Fructose ( mol/ejac) 34, , Male smoking, n (%) 4 (20) 121 (43.2).042 a a P based on 2 test. spermatids. Inhibin B levels did not contribute further to the accuracy of the regression model. The number of testicular tubuli with elongated spermatids was significantly associated with FSH concentrations (P.001) and with volume of the respective testis (P.001) but not with inhibin B levels (P.144) in multiple regression models. Including all 203 couples, noninvasive data were used to assess cutoff values concerning the ultimate outcome of fertility treatment: clinical pregnancy and live birth (Table 3). The significant result for FSH levels was confirmed by using receiver operating characteristics, resulting in 100% specificity for the prediction of no pregnancy (P.008) or no live birth (P.013) in the case of FSH levels of 20 IU/L (Fig. 3A and B). Cigarette smoking by men exerted a negative impact on treatment outcome: smokers were significantly underepresented in those couples who achieved clinical pregnancies (P.014) and live births (P.042; Table 3). Female smoking was rare within the total cohort (n 15). Pregnancies did not occur in these women in whom 20 TESE-ICSI treatments were performed. In the couples in which elongated spermatids were found, altogether 209 ICSI cycles from frozen testis samples were performed. These resulted in 23 pregnancies, leading to 20 live births of 26 children, including 6 twin pregnancies. Table 4 gives an overview of invasive parameters as predictors of various steps of ICSI. The degree of completely maintained spermatogenesis within the biopsy as expressed as percentage of all tubuli containing elongated spermatids remained the only invasive predictor of clinical pregnancies and live births in the simple approach using Mann- Whitney tests. This parameter was confirmed (P.02 and P.04, respectively) in backwards stepwise binomial regression models. The number of embryos transferred as major factor of influence of outcome in terms of clincial pregnancies and live births was included in the models (P.02 and P.04, respectively). Probability curves for the achievement of clinical pregnancies and live births relative to the degree of completely maintained spermatogenesis within the biopsy were created (Fig. 4), assuming a two-embryo transfer. 344 Zitzmann et al. Chances for fatherhood in azoospermia Vol. 86, No. 2, August 2006

7 FIGURE 3 Outcome of infertility treatment attempts (n 300) in relation to FSH levels. Inclusion of all patients and treatment cycles in regard to presence of sperm and success, hence, clinical pregnancy (A) or live birth (B). The cutoff level of 20 IU/L as threshold for positive outcome is indicated as thin horizontal line (receiver operating characteristics, with P.008 for clinical pregnancies and P.014 for live births). DISCUSSION Intracytoplasmic sperm injection with sperm retrieved from a testicular biopsy represents a treatment modality for azoospermic men who desire parenthood (4, 8 11, 23 25). However, as testis volume, serum FSH and inhibin B cannot be used absolutely to predict the presence of spermatozoa (3 5, 15, 26 28), we demonstrate regression models of high accuracy which can be used when counseling patients concerning the probability of a feasible ICSI procedure after surgery for TESE (Fig. 2). Nevertheless, useful predictions for high or low chances of sperm retrieval are only possible given extreme values (high testicular volume plus low FSH levels, or low testicular volume plus high FSH levels). In many cases (i.e., testicular volume of about 15 ml, FSH of about 15 IU/L), the patient can only be told that chances for sperm retrieval by TESE are within the range of flipping a coin. A most helpful indicator for successful paternity is the presurgical, continuous parameter of FSH levels. Our data suggest that although sperm can be found when FSH is 20 TABLE 4 Baseline comparisons of invasive parameters putatively influencing the outcome of 209 TESE-ICSI treatments in the 112 couples with azoospermic men from whom elongated sperm were retrieved. Parameter Fertilization of oocytes Clinical pregnancies Life births Success, n (%) 176 (84.2) 23 (11.0) 20 (9.6) P values a Volume of biopsied testis (ml).001 b Percentage of tubuli with elongated spermatids (%).001 b.01 b.03 b No. of motile sperm in the fresh biopsy material.009 b Grade A motility of sperm in fresh biopsy material (%).003 b No. of transferred embryos.02 b.04 b a Mann-Whitney test. Statistically significant, successful vs. nonsuccessful groups. Fertility and Sterility 345

8 FIGURE 4 Probability curves to predict chances of clinical pregnancies (dotted line) and live births (solid line) depending on the percentage of testicular tubuli containing elongated spermatids. Pregnancies are given as gray circles, twins are given as smaller black circles. IU/L (Fig. 1) and ICSI can be performed in these patients, a positive outcome in terms of clinical pregnancies or live births is likely only in patients who have relatively intact spermatogenesis, as reflected by the degree of completely maintained spermatogenesis within the biopsy (expressed as percentage of all tubuli containing elongated spermatids; Fig. 3A and B). Nevertheless, different conditions of ICSI and embryo culture in other countries influence the success rate, so that this threshold of FSH of 20 IU/L to live birth after TESE- ICSI can be passed in some cases (29). In agreement with previously published results, cigarette smoking is an adverse factor concerning the outcome of assisted reproduction techniques (30). The number of sperm present in the fresh biopsy material, as well as their motility, influences the success of initial in vitro embryogenesis. Nevertheless, the degree of completely maintained spermatogenesis within the biopsy, expressed as percentage of all tubuli containing elongated spermatids, is the only invasive parameter that is predictive for clinical pregnancies and also live births. This is also in agreement with the threshold of FSH levels observed and suggests that the degree of spermatogenesis reflects the ability of elongated spermatids to induce and maintain a pregnancy after TESE-ICSI, but FSH levels alone cannot predict ICSI success. Obviously, a certain number of sperm have to be formed to provide a significant number of germ cells with intact DNA. Previous comparisons between ejaculated sperm and those retrieved by TESE in regard to chromatin condensation have demonstrated that nuclear maturity was significantly higher in ejaculated sperm (31). Chromatin condensation during spermiogenesis is caused by histone-to-protamine exchange. Spermatozoa obtained by TESE exhibit a decrease in the protamine-1 mrna (PRM1) to protamine-2 mrna (PRM2) ratio with abating spermatogenic efficacy. This ratio also relates to success rates during fertilization (32, 33). It has to be mentioned that FSH levels do not measure the functional aspects of sperm and that such a functional assessment has to be performed by methods such as DNA fragmentation analysis, FISH, or TUNEL assays (e.g., as in Hammadeh et al. [31] and Steger et al. [32, 33]). Nevertheless, according to our results, FSH levels and the degree of completely maintained spermatogenesis reflect the ability of sperm to induce healthy embryo development. Thus, both entities may be considered as indirect indicators of sperm integrity. Our findings corroborate data demonstrating a higher success rate of TESE-ICSI in men with vasectomy compared with those with testicular disorders (34). The data also demonstrate that a categorization of azoospermic patients into those with obstructive and nonobstructive azoospermia often is not advisable because there is a marked variation within the determining variables. Figure 1 demonstrates that the clinical entity of azoospermia locates continuously within the spectrum of FSH levels and testicular volume. If there would be a clear distinction between obstruction and nonobstruction in all cases of azoospermia, the distribution of data points would show two markedly separated clusters, one in the low FSH high testicular volume area and one in the high FSH low testicular volume area; this is, however, not the case. As a consequence, our study demonstrates that the only useful predictor of sperm retrieval from a testicular biopsy and consecutive ICSI success is histopathology (9, 27), and it stresses the usefulness of FSH levels and testis volume measurement for presurgical counseling. We show that continuous parameters retrieved from assessment of testicular tissue samples are a valuable and precise tool for determining the approximate success rate of TESE-ICSI and, hence, for counseling patients more adequate than relying on the mere distinction between obstruction and nonobstruction. This is in agreement with the observation that the FSH concentrations are correlated positively with the number of tubuli containing only Sertoli cells (35, 36). In conclusion, the probability of live births after TESE- ICSI depends strongly on FSH levels as indicator for the degree of completely maintained spermatogenesis within the biopsy as well as the histology of TESE samples. Couples currently are being counseled in our institution on the basis of these data and appreciate the information, especially in cases with low probabilities. The couple s final choice to undergo therapy or seek alternative methods is, of course, 346 Zitzmann et al. Chances for fatherhood in azoospermia Vol. 86, No. 2, August 2006

9 also strongly influenced by individual attitudes toward hope and discomfort. Acknowledgment: The authors thank Susan Nieschlag, M.A., for language editing. REFERENCES 1. Tournaye H, Camus M, Goossens A, Liu J, Nagy P, Silber S, et al. Recent concepts in the management of infertility because of nonobstructive azoospermia. Hum Reprod 1995;10: Devroey P, Liu J, Nagy Z, Tournaye H, Silber SJ, van Steirteghem AC. Normal fertilization of human oocytes after testicular sperm extraction and intracytoplasmic sperm injection. Fertil Steril 1994;62: Devroey P, Liu J, Nagy Z, Goossens A, Tournaye H, Camus M, et al. Pregnancies after testicular sperm extraction and intracytoplasmic sperm injection in non-obstructive azoospermia. Hum Reprod 1995;10: Silber SJ, van Steirteghem A, Nagy Z, Liu J, Tournaye H, Devroey P. Normal pregnancies resulting from testicular sperm extraction and intracytoplasmic sperm injection for azoospermia due to maturation arrest. Fertil Steril 1996;66: Yemini M, Vanderzwalmen P, Mukaida T, Schoengold S, Birkenfeld A. Intracytoplasmic sperm injection, fertilization, and embryo transfer after retrieval of spermatozoa by testicular biopsy from an azoospermic male with testicular tubular atrophy. Fertil Steril 1995;63: Kahraman S, Ozgur S, Alatas C, Aksoy S, Balaban B, Evrenkaya T, et al. High implantation and pregnancy rates with testicular sperm extraction and intracytoplasmic sperm injection in obstructive and non-obstructive azoospermia. Hum Reprod 1996;11: Fahmy I, Mansour R, Aboulghar M, Serour G, Kamal A, Tawab NA, et al. Intracytoplasmic sperm injection using surgically retrieved epididymal and testicular spermatozoa in cases of obstructive and nonobstructive azoospermia. Int J Androl 1997;20: Mansour RT, Kamal A, Fahmy I, Tawab N, Serour GI, Aboulghar MA. Intracytoplasmic sperm injection in obstructive and non-obstructive azoospermia. Hum Reprod 1997;2: Vernaeve V, Tournaye H, Osmanagaoglu K, Verheyen G, van Steirteghem A, Devroey P. Intracytoplasmic sperm injection with testicular spermatozoa is less successful in men with non-obstructive azoospermia than in men with obstructive azoospermia. Fertil Steril 2003;79: Palermo GD, Schlegel PN, Hariprashad JJ, Ergun B, Mielnik A, Zaninovic N, et al. Fertilization and pregnancy outcome with intracytoplasmic sperm injection for azoospermic men. Hum Reprod 1999;14: De Croo I, Van der Elst J, Everaert K, De Sutter P, Dhont M. Fertilization, pregnancy and embryo implantation rates after ICSI in cases of obstructive and non-obstructive azoospermia. Hum Reprod 2000;15: Madgar I, Hourvitz A, Levron J, Seidman DS, Shulman A, Raviv GG, et al. Outcome of in vitro fertilization and intracytoplasmic injection of epididymal and testicular sperm extracted from patients with obstructive and non-obstructive azoospermia. Fertil Steril 1998;69: Luetjens CM, Gromoll J, Engelhardt M, von Eckardstein S, Bergmann M, Nieschlag E, et al. Manifestation of Y-chromosomal deletions in the human testis: a morphometrical and immunohistochemical evaluation. Hum Reprod 2002;17: Raleigh D, O Donnell L, Southwick GJ, de Kretser DM, McLachlan RI. Stereological analysis of the human testis after vasectomy indicates impairment of spermatogenic efficiency with increasing obstructive interval. Fertil Steril 2004;81: von Eckardstein S, Simoni M, Bergmann M, Weinbauer GF, Gassner P, Schepers AG, et al. Serum inhibin B in combination with serum follicle-stimulating hormone (FSH) is a more sensitive marker than serum FSH alone for impaired spermatogenesis in men, but cannot predict the presence of sperm in testicular tissue samples. J Clin Endocrinol Metab 1999;84: Behre HM, Yeung CH, Holstein AF, Weinbauer GF, Gassner P, Nieschlag E. Diagnosis of male infertility and hypogonadism. In: Nieschlag E, Behre HM, eds. Andrology: male reproductive health and dysfunction. Heidelberg, Germany: Springer, 2000: World Health Organization. WHO laboratory manual for the examination of human semen and sperm-cervical mucus interaction. 4th ed. Cambridge, United Kingdom: Cambridge University Press, 1999: Cooper TG, Weidner W, Nieschlag E. The influence of inflammation of the human male genital tract on secretion of the seminal markers alpha-glucosidase, gycerophosphocholine, carnitine, fructose and citric acid. Int J Androl 1990;13: Cooper TG, Yeung CH, Nashan D, Jockenhövel F, Nieschlag E. Improvement in the assessment of human epididymal function by the use of inhibitors in the assay of alpha-glucosidase in seminal plasma. Int J Androl 1990;13: Cooper TG, Jockenhövel F, Nieschlag E. Variation in seminal parameters from fathers. Hum Reprod 1991;6: Palermo GD, Joris H, Devroey P, Van Steirteghem AC. Pregnancies after intracytoplasmic injection of single spermatozoon into an oocyte. Lancet 1992;340: van Steirteghem AC, Nagy Z, Joris H, Liu J, Staessen C, Smitz J, et al. High fertilization and implantation rates after intracytoplasmic sperm injection. Hum Reprod 1993;8: Mercan R, Urman B, Alatas C, Aksoy S, Nuhoglu A, Isiklar A, et al. Outcome of testicular sperm retrieval procedures in non-obstructive azoospermia: percutaneous aspiration versus open biopsy. Hum Reprod 2000;15: Balaban B, Urman B, Isiklar A, Alatas C, Mercan R, Aksoy S, et al. Blastocyst transfer following intracytoplasmic injection of ejaculated, epididymal or testicular spermatozoa. Hum Reprod 2001;16: Bukulmez O, Yucel A, Yarali H, Bildirici I, Gurgan T. The origin of spermatozoa does not affect intracytoplasmic sperm injection outcome. Eur J Obstet Gynecol Reprod Biol 2001;94: Lewin A, Weiss DB, Friedler S, Ben-Shachar I, Porat-Katz A, Meirow D, et al. Delivery following intracytoplasmic injection of mature sperm cells recovered by testicular fine needle aspiration in a case of hypergonadotropic azoospermia due to maturation arrest. Hum Reprod 1996; 11: Tournaye H, Verheyen G, Nagy P, Ubaldi F, Goossens A, Silber S, et al. Are there any predictive factors for successful testicular sperm recovery in azoospermic patients? Hum Reprod 1997;12: Ezeh UI, Moore HD, Cooke ID. Correlation of testicular sperm extraction with morphological, biophysical and endocrine profiles in men with azoospermia due to primary gonadal failure. Hum Reprod 1998; 13: Hibi H, Ohori T, Yamada Y, Honda N, Asada Y. Probability of sperm recovery in non-obstructive azoospermic patients presenting with testes volume less than 10 ml/fsh level exceeding 20 miu/ml. Arch Androl 2005;51: Zitzmann M, Rolf C, Nordhoff V, Schrader G, Rickert-Föhring M, Gassner P, et al. Male smokers have a decreased success rate for in vitro fertilization and intracytoplasmic sperm injection. Fertil Steril 2003;79: Hammadeh ME, al-hasani S, Doerr S, Stieber M, Rosenbaum P, Schmidt W, et al. Comparison between chromatin condensation and morphology from testis biopsy extracted and ejaculated spermatozoa and their relationship to ICSI outcome. Hum Reprod 1999;14: Steger K, Failing K, Klonisch T, Behre HM, Manning M, Weidner W, et al. Round spermatids from infertile men exhibit decreased protamine-1 and -2 mrna. Hum Reprod 2001;16: Steger K, Fink L, Failing K, Bohle RM, Kliesch S, Weidner W, et al. Decreased protamine-1 transcript levels in testes from infertile men. Mol Hum Reprod 2003;9: Pasqualotto FF, Rossi LM, Guilherme P, Ortiz V, Iaconelli A Jr, Borges E Jr. Etiology-specific outcomes of intracytoplasmic sperm injection in azoospermic patients. Fertil Steril 2005;83: Bergmann M, Behre HM, Nieschlag E. Serum FSH and testicular morphology in male infertility. Clin Endocrinol (Oxf) 1994;40: Leifke E, Simoni M, Kamischke A, Gromoll J, Bergmann M, Nieschlag E. Does the gonadotrophic axis play a role in the pathogenesis of Sertoli-cell-only syndrome? Int J Androl 1997;20: Fertility and Sterility 347