Serum inhibin B cannot predict testicular sperm retrieval in patients with non-obstructive azoospermia

V. Vernaeve,1, H. Tournaye, J. Schiettecatte, G. Verheyen, A.Van Steirteghem and P. Devroey

Centre For Reproductive Medicine, University Hospital, Dutch-speaking Brussels Free University (Vrije Universiteit Brussel), Laarbeeklaan, Belgium


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
BACKGROUND: Serum inhibin B, a direct product of the Sertoli cells, may serve as a marker of spermatogenesis. The present retrospective study aimed at evaluating the predictive value of inhibin B for retrieving testicular sperm in non-obstructive azoospermic men. METHODS: The serum inhibin B concentration before sperm retrieval was reviewed in 185 non-obstructive azoospermic patients. RESULTS: Testicular sperm were successfully recovered in 92 of 185 patients (49.7%). The mean inhibin B concentration in these patients was 37.3 pg/ml. No sperm were found in 93 patients (50.3%), and the mean serum inhibin B concentration was 44.9 pg/ml. The discrimination between successful and unsuccessful sperm retrieval was analysed using the receiver operating characteristics (ROC) curve analysis. The best discriminating inhibin B concentration was 13.7 pg/ml (sensitivity 44.6%, specificity 63.4%) with an area under the ROC curve (AUC) of 0.51. Combining both serum FSH and inhibin B did not improve the predictive value: the AUC of inhibin B in men with a serum FSH concentration <25 and 25 IU/l (being the best threshold value in the population studied) was respectively 0.53 and 0.50. The AUC of the inhibin B:FSH ratios was 0.55. CONCLUSIONS: This analysis shows that inhibin B, either alone or in combination with serum FSH, fails to predict the presence of sperm in men with non-obstructive azoospermia undergoing testicular sperm extraction.

Key words: ICSI/inhibin B/male infertility/non-obstructive azoospermia/TESE


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The use of testicular sperm extraction (TESE) to recover testicular sperm for ICSI has created new possibilities for the treatment of men with non-obstructive azoospermia (NOA). However, testicular sperm recovery is successful in only 50% of cases (Tournaye et al., 1995Go). An unsuccessful sperm recovery procedure also has important emotional and financial implications, as well as complications such as devascularization and fibrosis (Friedler et al., 1997Go; Schlegel and Su, 1997Go; Ron-El et al., 1998Go). Objective counselling based on predictive factors is therefore important. Different parameters such as testicular size, serum FSH level and preliminary testicular histology have previously been investigated (Tournaye et al., 1997Go; Ezeh et al., 1999Go). Only the latter parameter appeared to have any clinical value in predicting sperm during TESE (Tournaye et al., 1997Go; Ezeh et al., 1999Go). However, the absence of sperm in one single testicular biopsy does not preclude the presence of some sperm in the testes (Gottschalk-Sabag et al., 1995Go).

Serum inhibin is a glycoprotein of gonadal origin which has an inhibitory effect on gonadotrophin secretion. It is known to originate from the Sertoli cells, and its secretion is regulated by interaction with germ cells (Pineau et al., 1990Go). This glycoprotein is a heterodimer consisting of an {alpha}-subunit and one of two different ß-subunits: ßA (inhibin A) and ßB (inhibin B). Inhibin B is the most important form in human males and may serve as a marker of spermatogenesis (Anawalt et al., 1996Go; Illingworth et al., 1996Go; Jensen et al., 1997Go; Kolb et al., 2000Go). The normal range of serum concentration in proven fathers is 94 to 327 pg/ml (Von Eckardstein et al., 1999Go).

One report (Ballesca et al., 2000Go) suggested that inhibin B might also be a good non-invasive predictor for retrieving sperm by TESE, though another study failed to corroborate this finding (Von Eckardstein et al., 1999Go). The aim of the present study was to assess the predictive value of inhibin B before testicular sperm retrieval in a large population of patients with NOA.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Patients
Between January 1996 and December 1999, blood samples were taken on the morning before surgery from all patients undergoing a TESE procedure; the serum was separated and stored at –20°C. NOA patients were selected according to testicular histology. This group included those patients not showing normal spermatogenesis or hypospermatogenesis, including patients with complete or incomplete maturation arrest, complete or incomplete germ-cell aplasia (Sertoli cell-only) and tubular sclerosis and atrophy. All patients had absolute azoospermia (no sperm found in any of the semen analyses after centrifugation at high speed). Klinefelter patients and patients with hypogonadotrophic hypogonadism were excluded. A total of 185 patients was selected for further study.

Hormone analyses
Serum inhibin B was measured using a solid-phase sandwich enzyme-linked immunosorbent assay (ELISA) (Inhibin B, Oxford Bio-Innovation Ltd, Oxford, UK). The samples were pre-treated with detergent (sodium dodecyl sulphate), heated to 100°C, and exposed to hydrogen peroxide. Pre-treated samples were incubated in microtitre plates, coated with a monoclonal antibody to the ßB-subunit of inhibin. The detection antibody was a monoclonal antibody specific for the {alpha}-subunit of inhibin and coupled to alkaline phosphatase. The kit was calibrated against recombinant 32 kDa inhibin B. The detection limit of the assay was 8 pg/ml, and the between-assay coefficient of variation was 7.5 at 197 pg/ml (n = 8). A cross-reactivity of <1% with inhibin A was apparent, but no cross-reactivity was found with pro-alpha C subunits or activins. Patients with inhibin B concentrations below the detection level were included in the analysis and assigned a value of 8 pg/ml.

FSH assessment was performed during the initial work-up for each man before embarking on a TESE procedure. Serum FSH was measured using a commercially available monoclonal immunoradiometric assay (bioMérieux, France). The FSH assay, calibrated against the 2nd IRP 78/549, had an analytical sensitivity of 0.25 IU/l and within- and between-run coefficients of variation of<4 and <9% respectively.

Testicular sperm recovery
Open excisional testicular biopsies were taken under general anaesthesia (Tournaye et al., 1997Go) on the day of oocyte retrieval. Surgery was discontinued when sperm or late elongated spermatids (stage Sd2) (Clermont, 1963Go) were found on wet preparations of the biopsies, or when a representative number of samples (at least four biopsies on each side) were taken randomly. Wet preparation of testicular tissue was performed by mechanical shredding as described previously (Jow et al., 1993Go; Verheyen et al., 1995Go). Since 1998, enzymatic digestion of the testicular tissue with collagenase type IV has been performed if no sperm were found after mechanical shredding (Crabbé et al., 1998Go). During surgery, a randomly taken single small biopsy was sent for histopathological examination.

Statistical analysis
The two groups of patients with either successful or unsuccessful sperm recovery were evaluated by receiver operating characteristics (ROC) analysis (Zweig and Campbell, 1993Go) using Medcalc (Medcalc Software, Ghent, Belgium). ROC graphs are plots of all the sensitivity (y-axis) and specificity (x-axis) pairs which are possible for all levels of a particular parameter, for example serum inhibin B concentration. Calculation of the area under the curve provides a quantitative measure of accuracy, namely the ability of this parameter to discriminate between two conditions. A test with perfect discrimination has an ROC plot that passes through the upper left corner, where the true positive fraction is 1.0, or 100% (perfect sensitivity), and the false-positive fraction is 0 (perfect specificity). The theoretical plot for a test with no discrimination (identical distributions of results for the two groups) is a 45° diagonal line from the lower left corner (0% true-positive rate and 0% false-positive rate) to the upper right corner (100% true-positive rate and 100% false-positive rate) with an area under the curve of 0.5 (Zweig and Campbell, 1993Go).


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The mean (± SD) age of the 185 azoospermic patients was 35.6 ± 7.3 (range 21–58) years. Among 185 patients, 128 (69.2%) showed complete or incomplete germ-cell aplasia (Sertoli cell-only), 49 (26.5%) showed complete or incomplete maturation arrest, and eight (4.3%) showed tubular sclerosis and atrophy. Overall, testicular sperm were successfully recovered in 92 of the 185 patients (49.7%). No sperm were found in 93 patients (50.3%). Sperm recovery in the different subpopulations is listed in Table IGo. As expected, a recovery rate of ~50% was found in the two largest subgroups of germ-cell aplasia and maturation arrest.


View this table:
[in this window]
[in a new window]
 
Table I. Successful and unsuccessful sperm recovery rates in the histological subpopulations
 
The mean (± SEM) serum inhibin B concentration in the study population was 41.2 ± 4.7 pg/ml; concentrations in the subgroups of germ-cell aplasia, maturation arrest and tubular sclerosis and atrophy were 23.5 ± 2.8, 92.6 ± 14.0 and 8.0 ± 0.0 pg/ml respectively. The mean (± SEM) inhibin B concentration in patients with sperm was 37.3 ± 5.4 pg/ml, and 44.9 ± 7.7 pg/ml where no sperm were found. The distribution of inhibin B concentrations among the study population and in different subgroups is shown in Figure 1Go. Clearly, there was a wide distribution of inhibin B concentrations, and inhibin B was essentially undetectable (<8 pg/ml) in a large number of patients.



View larger version (12K):
[in this window]
[in a new window]
 
Figure 1. Distribution of serum inhibin B concentrations in patients with and without successful testicular sperm recovery in the study population and in subgroups of germ-cell aplasia, maturation arrest and tubular sclerosis and atrophy. Sperm + = successful sperm retrieval; sperm – = unsuccessful sperm retrieval.

 
The predictive value of inhibin B for discriminating between successful and failed TESE was assessed by calculating the area under the ROC curve. The area under the curve in the present population was 0.51. The best inhibin B concentration for discriminating between successful and failed TESE was 13.7 pg/ml (sensitivity 44.6%, specificity 63.4%) (Figure 2Go). The areas under the ROC curves in the aplasia and arrest subgroups were 0.6 and 0.7 respectively, with the best discriminative concentration of 13.7 pg/ml (sensitivity 39.7%, specificity 75.4%) and 72.4 pg/ml respectively (sensitivity 73.1%, specificity 63.6%) (Figures 3 and 4GoGo).



View larger version (31K):
[in this window]
[in a new window]
 
Figure 2. Receiver operating characteristics (ROC) curve of inhibin B concentrations for discriminating between successful and failed testicular sperm extraction in all non-obstructive azoospermia patients. The best discriminating concentration is indicated(13.7 pg/ml).

 


View larger version (31K):
[in this window]
[in a new window]
 
Figure 3. Receiver operating characteristics (ROC) curve of inhibin B concentrations for discriminating successful and failed testicular sperm extraction in patients with germ-cell aplasia. The best discriminating concentration is indicated (13.7 pg/ml).

 


View larger version (31K):
[in this window]
[in a new window]
 
Figure 4. Receiver operating characteristics (ROC) curve of inhibin B concentrations for discriminating successful and failed testicular sperm extraction in patients with maturation arrest. The best discriminating concentration is indicated (72.4 pg/ml).

 
The best serum FSH concentration for discriminating between successful and failed TESE was 25 IU/l (sensitivity 74.3%, specificity 44.3%) with an area under the curve of 0.56. On the basis of this best threshold, the study population was subdivided into two categories: men with serum FSH <25 IU/l, and men with FSH >=25 IU/l. The area under the ROC curve of inhibin B in men with serum FSH <25 IU/l was 0.53 (sensitivity 97.9%, specificity 13.2%) (Figure 5Go, left panel), while that in men with serum FSH >=25 IU/l was 0.50 (sensitivity 100%, specificity 6.2%) (Figure 5Go, right panel). The area under the ROC curve of the inhibin B:FSH ratios was 0.55 (sensitivity 75.7%, specificity 41.4%) (Figure 6Go).




View larger version (56K):
[in this window]
[in a new window]
 
Figure 5. Receiver operating characteristics (ROC) curve of inhibin B concentrations for discriminating successful and failed testicular sperm extraction in non-obstructive azoospermia patients with a serum FSH concentration <25 IU/l (left panel, the best discriminating concentration is indicated at 213.8 pg/ml) and with an FSH concentration of >=25 IU/l (right panel, the best discriminating concentration is indicated at 34.3 pg/ml).

 


View larger version (29K):
[in this window]
[in a new window]
 
Figure 6. Receiver operating characteristics (ROC) curve of inhibin B:FSH ratios for discriminating successful and failed testicular sperm extraction in non-obstructive azoospermia patients. The best discriminating value is indicated (0.32).

 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The recovery of testicular sperm for ICSI has become a useful therapeutic approach in NOA patients (Devroey et al., 1995Go; Tournaye et al., 1995Go, 1996Go). This procedure, however, may have side effects (Friedler et al., 1997Go; Schlegel and Su, 1997Go; Ron-El et al., 1998Go) and is not always successful in terms of finding sperm that can be used for ICSI (Tournaye et al., 1995Go). An objective counselling based on predictive parameters is therefore needed. Testicular size, serum FSH level and testicular histology have been investigated, but only the presence of sperm in a preliminary biopsy appeared to have some predictive power (Tournaye et al., 1997Go; Ezeh et al., 1999Go). However, even the absence of sperm in a single testicular biopsy does not guarantee a complete lack of sperm in the testes (Gottschalk-Sabag et al., 1995Go; Tournaye et al., 1995Go). Additional predictive parameters are therefore more than welcome. Serum inhibin B, a direct product of the Sertoli cells, is a marker of spermatogenesis (Anawalt et al., 1996Go; Illingworth et al., 1996Go; Jensen et al., 1997Go; Kolb et al., 2000Go). However, controversy persists regarding the role of inhibin B as a predictor for finding sperm by TESE. In two small studies, inhibin B had a clear predictive value for finding sperm by surgical testicular sperm retrieval (Ballesca et al., 2000Go) or by diagnostic testicular biopsy (Pierik et al., 1998Go). Others (Von Eckardstein et al., 1999Go) found an improvement in sensitivity in the prediction for finding sperm on histology or in wet preparations, but did not find inhibin B concentration accurate enough to predict TESE. In the present study, discrimination between successful and unsuccessful sperm retrieval was assessed using the ROC curve analysis. In the present study population, with an area under the curve of 0.51, the situation was close to having a theoretical plot for a test, with no discrimination (Figure 2Go). The best inhibin B concentration for discriminating between successful and failed TESE was 13.7 pg/ml (sensitivity 44.6%, specificity 63.4%).

As shown in Figure 1Go, there was also a wide distribution of inhibin B concentrations, with an important overlap between successful and unsuccessful sperm retrievals. Many patients also had an inhibin B concentration below the detectable limit.

Despite the fact that the role of inhibin B as a marker for spermatogenesis has been established, the inability to detect inhibin B concentrations <8 pg/ml with the currently available assay might be a cause of failure to predict sperm retrieval before TESE.

Inhibin B has been reported to be associated with late stages of spermatogenesis (Pineau et al., 1990Go; Andersson et al., 1998Go). Because only a few men in the present study population may have had spermatogenesis progressing beyond the spermatocyte stage in a representative number of their seminiferous tubules, inhibin B might have a limited predictive power.

The difference with the above-mentioned studies may have been be due to the number of patients included, and also to differences in patient selection. Prediction based on inhibin B in men appears to be better in those with a higher spermatogenetic output (Figure 4Go). In the present study, 73% of azoospermic patients had severely disturbed spermatogenesis (germ-cell aplasia and tubular sclerosis and atrophy), whereas in studies reported earlier (Pierik et al., 1998Go; Von Eckardstein et al., 1999Go; Ballesca et al., 2000Go) it was mainly men with a better spermatogenetic function who were included.

The difference in endpoint in the different studies might also be a reason for the discrepancy in results. In one study (Pierik et al., 1998Go) a quantitative endpoint (the Johnsen score) was used, while others (Von Eckardstein et al., 1999Go) used the histological subtypes (semi-quantitative). In the present study and one other study (Ballesca et al., 2000Go), sperm retrieval was used as an endpoint (qualitative). As for the other previously examined predictors, namely testicular size, serum FSH level and testicular histology, the poor correlation of successful sperm retrieval with inhibin B might also have been due to the non-quantitative nature of this endpoint. Standardization of sperm retrieval is difficult in a clinical setting, and consequently the amount of testicular tissue was not uniform for all patients. However, an attempt was made to standardize this endpoint as well as possible by taking at least four biopsies from each side if no sperm were directly found at wet preparation. These biopsies were taken randomly because of the heterogeneously distributed spermatogenesis throughout the testis (Gottschalk-Sabag et al., 1995Go; Tournaye et al., 1996Go).

The predictive power of assessing inhibin B combined with serum FSH level has also been assessed (Von Eckardstein et al., 1999Go). Although the combination showed a better predictive value, these authors nevertheless concluded that inhibin B alone or in combination with FSH was of limited predictive value for TESE. The present results also indicate that combining FSH and inhibin B, either by using a ratio or by creating subgroups of patients on the basis of their serum FSH levels, does not improve the predictive power of inhibin B. Although the role of inhibin B is clear as a marker for spermatogenesis, the present data indicate that it does not have any role in predicting the presence of sperm before TESE.

In the future, other markers—both genetical and/or biochemical—will likely be investigated in order to improve the prediction of successful TESE. One study (Brandell et al., 1998Go) has outlined results from a limited series of patients in whom the presence of AZFb micro-deletions of the Y chromosome indicated an unsuccessful TESE.


    Acknowledgements
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The authors thank Walter Meul for technical assistance, Yvo Pletinckx of the andrology laboratory, and Frank Winter of the Language Education Centre of our University for editing this manuscript.


    Notes
 
1 To whom correspondence should be addressed at: Centre For Reproductive Medicine, University Hospital, Dutch-speaking Brussels Free University (Vrije Universiteit Brussel), Laarbeeklaan 101, B-1090, Belgium. E-mail: vvernaeve{at}skynet.be Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Andersson, A., Muller, J. and Skakkebaek, N.E. (1998) Different roles of prepubertal and postpubertal germ cells and Sertoli cells in the regulation of serum inhibin B levels. J. Clin. Endocrinol. Metab., 83, 4451–4458.[Abstract/Free Full Text]

Anawalt, B.D., Bebb, R.A., Matsumoto, Groome, N.P., Illingworth P.J., McNeilly, A.S. and Bremner W.J. (1996) Serum inhibin B levels reflect Sertoli cell function in normal men and men with testicular dysfunction. J. Clin. Endocrinol. Metab., 81, 3341–3345.[Abstract]

Ballesca, J.L., Balasch, J., Calafell, J.M., Alvarez, R., Fabregues, F., Martinez de Osaba, M.J., Ascaso, C. and Vanrell, J.A. (2000) Serum inhibin B determination is predictive of successful testicular sperm extraction in men with non-obstructive azoospermia. Hum. Reprod., 15, 1734–1738.[Abstract/Free Full Text]

Brandell, R.A., Mielnik, A., Liotta, D., Ye, Z., Veeck, L.L., Palermo, G.D. and Schlegel, P.N. (1998) AZFb deletions predict the absence of sperm with testicular sperm extraction: preliminary report of a prognostic genetic test. Hum. Reprod., 13, 2812–2815.[Abstract/Free Full Text]

Clermont, Y. (1963) The cycle of the seminiferous epithelium in men. Am. J. Anat., 112, 35–52.[ISI]

Crabbé, E., Verheyen, G., Silber, S., Tournaye, H., Van de Velde, H., Goossens, A. and Van Steirteghem, A. (1998) Enzymatic digestion of testicular tissue may rescue the intracytoplasmic sperm injection cycle in some patients with non-obstructive azoospermia. Hum. Reprod., 13, 2791–2796.[Abstract/Free Full Text]

Devroey, P., Liu, J., Nagy, Z., Goossens, A., Tournaye, H., Camus, M., Van Steirteghem, A. and Silber, S. (1995) Pregnancies after testicular sperm extraction and intracytoplasmic sperm injection in non-obstructive azoospermia. Hum. Reprod., 10, 1457–1460.[Abstract]

Ezeh, U.I.O., Taub, N.A., Moore, H.D.M. and Cooke, I.D. (1999) Establishment of predictive variables associated with testicular sperm retrieval in men with non-obstructive azoospermia. Hum. Reprod., 14, 1005–1012.[Abstract/Free Full Text]

Friedler, S., Raziel, A., Strassburger, D., Soffer, Y., Komarovsky, D. and Ron-El, R. (1997) Testicular sperm retrieval by percutaneous fine needle sperm aspiration compared with testicular sperm extraction by open biopsy in men with non-obstructive azoospermia. Hum. Reprod., 12, 1488–1493.[Abstract]

Gottschalk-Sabag, S., Weiss, D.B., Folb-Zacharow, N. and Zukerman, Z. (1995) Is one testicular specimen sufficient for quantitative evaluation of spermatogenesis? Fertil. Steril., 64, 399–402.[ISI][Medline]

Illingworth, P.J., Groome, N.P., Byrd, W., Rainey, W.E., McNeilly, A.S., Mather, J.P. and Bremner, W.J. (1996) Inhibin-B: a likely candidate for the physiologically important form of inhibin in men. J. Clin. Endocrinol. Metab., 81, 1321–1325.[Abstract]

Jensen, T.K., Andersson, A.-M., Hjollund, N.H.I., Scheike, T., Kolstad, H., Giwercman, A., Henriksen, T.B., Ernst, E., Bonde, J.P., Olsen, J. et al. (1997) Inhibin B as a serum marker of spermatogenesis: correlation to differences in sperm concentration and follicle-stimulating hormone levels. A study of 349 Danish men. J. Clin. Endocrinol. Metab., 82, 4059–4063.[Abstract/Free Full Text]

Jow, W.W., Steckel, J., Schlegel, P., Magid, M.S. and Goldstein, M. (1993) Motile sperm in human testis biopsy specimens. J. Androl., 14, 194–198.[Abstract]

Kolb, B.A., Stanczyk, F.Z. and Sokol, R.Z. (2000) Serum inhibin B levels in males with gonadal dysfunction. Fertil. Steril., 74, 234–238.[ISI][Medline]

Pierik, F.H., Vreeburg, J.T.M., Stijnen, T., De Jong, F.H. and Weber, R.F.A. (1998) Serum inhibin B as a marker of spermatogenesis. J. Clin. Endocrinol. Metab., 83, 3110–3114.[Abstract/Free Full Text]

Pineau, C., Sharpe, R.M., Saunders, P.T.K., Gerard, N. and Jegou, B. (1990) Regulation of Sertoli cell inhibin production and of inhibin {alpha}-subunit mRNA levels by specific germ cell types. Mol. Cell. Endocrinol., 72, 13–22.[ISI][Medline]

Ron-El, R., Strauss, S., Friedler, S., Strassburger, D., Komarovsky, D. and Raziel, A. (1998) Serial sonography and colour flow Doppler imaging following testicular and epididymal sperm extraction. Hum. Reprod., 13, 3390–3393.[Abstract]

Schlegel, P.N. and Su, L. (1997) Physiological consequences of testicular sperm extraction. Hum. Reprod., 12, 1688–1692.[Abstract]

Tournaye, H., Camus, M., Goossens, A., Nagy, Z., Silber, S., Van Steirteghem, A.C. and Devroey, P. (1995) Recent concepts in the management of infertility because of non-obstructive azoospermia. Hum. Reprod., 10 (Suppl. 1), 115–119.[ISI][Medline]

Tournaye, H., Liu, J., Nagy, P.Z., Camus, M., Goossens, A., Silber, S., Van Steirteghem, A.C. and Devroey, P. (1996) Correlation between testicular histology and outcome after intracytoplasmic sperm injection using testicular sperm. Hum. Reprod., 11, 127–132.[Abstract]

Tournaye, H., Verheyen, G., Nagy, P., Goossens, A., Ubaldi, F., Silber, S., Van Steirteghem, A. and Devroey, P. (1997) Are there any predictive factors for successful testicular sperm recovery in azoospermic patients? Hum. Reprod., 12, 80–86.[ISI][Medline]

Verheyen, G., De Croo, I., Tournaye, H., Pletinckx, I., Devroey, P. and Van Steirteghem, A. (1995) Comparison of four mechanical methods to retrieve sperm from testicular tissue. Hum. Reprod., 10, 2956–2959.[Abstract]

Von Eckardstein, S., Simoni, M., Bergmann, M., Weinbauer, G.F., Gassner, P., Schepers, A.G. and Nieschlag, E. (1999) 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., 84, 2496–2501.[Abstract/Free Full Text]

Zweig, M.H. and Campbell, G. (1993) Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clin. Chem., 39, 561–577.[Abstract/Free Full Text]

Submitted on April 24, 2001; resubmitted on October 2, 2001; accepted on December 5, 2001.