Archivio Italiano di Urologia e Andrologia 2018; 90, 2136 ORIGINAL PAPER Predictive factors of successful salvage microdissection testicular sperm extraction (mTESE) after failed mTESE in patients with non-obstructive azoospermia: Long-term experience at a single institute Cem Yücel, Salih Budak, Mehmet Zeynel Keskin, Erdem Kisa, Zafer Kozacioglu Department of Urology, Tepecik Training and Research Hospital, Izmir, Turkey. Objective: To observe the clinical practice of salvage microdissection testicular sperm extraction (mTESE) in patients with non-obstructive azoosper- mia (NOA) and to determine the factors that may predict the presence of spermatozoa in preoperative salvage mTESE. Methods: We retrospectively reviewed the medical records of 445 patients with the diagnosis of NOA, who had undergone the mTESE operation consecutively in our institution between the dates of March 2008 and June 2017. The study included a total of 49 patients with failure to detect spermatozoa in the first mTESE and who had then undergone salvage mTESE. In order to investigate the factors that predict the result of sal- vage mTESE, the patients were classified into two groups according to the outcome of salvage mTESE, as those with and without spermatozoa retrieval. Patients in these two groups were compared with regard to age, body mass index, history of varicocele, history of cryptorchidism, duration of infertility, outcomes of genetic analysis, results of hormone profiles and the testicular histopathology results of the first mTESE. Results: The sperm retrieval rate following salvage mTESE was observed to be 42.8%. Statistically a significant difference was determined between the mean follicle stimulating hormone (FSH) values of the groups (p = 0.013). No significant differ- ence was observed between the groups with regard to the remaining parameters. Conclusion: It was observed that among the factors that predict the success of sperm retrieval in salvage mTESE in patients with NOA and previous unsuccessful sperm retrieval in mTESE operation, only the pre-operative FSH level was observed to significantly correlate with the success in salvage mTESE. KEY WORDS: Azoospermia; Salvage; Testicular sperm extraction; Spermatozoa. Submitted 11 January 2018; Accepted 20 February 2108 Summary No conflict of interest declared. The current treatment of NOA is sperm retrieval from the testes via testicular sperm extraction (TESE) and using these sperms in intracytoplasmic sperm injection (ICSI) to obtain a healthy pregnancy (2). Currently, TESE opera- tions are performed in the guidance of a microscope (mTESE). Although it was initially reported that retrieval rate following a first TESE attempt in a well-defined NOA population was around 50%, recovery rates report- ed subsequently in literature were inconsistent (3). Unsuccessful sperm retrieval from the first TESE opera- tion results in negative emotional and financial effects. Salvage TESE offers a further chance of pregnancy for whom the first TESE has been unsuccessful. There is no clinical finding or test that precisely predicts the out- come of TESE preoperatively. Knowing the clinical char- acteristics that help predicting the outcomes of salvage TESE would be facilitative for the preoperative counsel- ing and clinical management of patients undergoing sal- vage TESE. Currently, there are not many studies that have published the clinical application of salvage TESE in the literature. The aim of this study was to observe the clinical practice of salvage mTESE in patients with NOA and to deter- mine the factors that may predict the presence of sper- matozoa in preoperative salvage mTESE, and help physi- cians determine the best candidates for this procedure. MATERIALS AND METHODS Patients and study design We retrospectively reviewed the medical records of 445 patients with the diagnosis of NOA, who had undergone the mTESE operation in our institution between the dates of March 2008 and June 2017. The diagnosis of NOA was confirmed by clinical findings, medical history, physical examination, serum hormone levels, genetic analysis and as suggested by the WHO guideline, 2 semen analysis. Semen analyses were obtained by masturbation after 3-4 days of sexual abstinence. The levels of serum total testosterone, follicular stimulat- ing hormone (FSH), luteinizing hormone (LH), estradi- ol, prolactin of the patients and the genetic analyses DOI: 10.4081/aiua.2018.2.136 INTRODUCTION Azoospermia is described as the absence of spermatozoa in the ejaculate and is observed in 1% of all men and 10- 15% of those with the complaint of infertility (1). Azoospermia is examined in two groups according to its etiology as obstructive and non-obstructive azoosper- mia. Non-obstructive azoospermia (NOA) is accepted as the absence of spermatozoa in the ejaculate due to min- imally developed or unproduced cells in the testicles. Yucel2_Stesura Seveso 28/06/18 16:40 Pagina 136 137Archivio Italiano di Urologia e Andrologia 2018; 90, 2 Predictive factors of successful salvage mTESE after failed mTESE (karyotype analysis and Y chromosome micro-deletion analysis) results were evaluated. For the hormone pro- file, blood was drawn from the antecubital vein of the patients after at least 8 hours of fasting. The micro-parti- cle enzyme immuno-assay method (Roche/Hitachi, Cobas e601, Indianapolis, IN, USA) was used to determine all the hormone levels. For the chromosome analysis, the peripheral venous blood samples of the patients were subjected to 72 hours of phytohemagglutinin-induced cell culture. The study included a total of 49 patients with NOA, who had previously mTESE and no sperm retrieval could be available. All patients included in the study had under- gone TESE operations according to the mTESE proce- dure both in the previous attempt and in the second attempt. In order to ascertain wound healing following the first operation, salvage TESE was planned for a min- imum of 3 months afterwards. No hormone therapy was administered to the patients in the time between the first TESE and salvage TESE. Those with a Y chromosome micro-deletion analysis revealing AZFa or AZFb, history of malignancy, those who were morbidly obese, those who had obstruction-related azoospermia, those who had undergone mTESE prior to the second ICSI attempt despite sperm retrieval being possible in the first TESE, and undergone multiple mTESEs, and those whose first TESE operation was conventional, were excluded from the study. The patients were classified into two groups as those with or without sperm retrieval in order to investi- gate the factors that predict the outcome of salvage mTESE. Patients in these two groups were compared with regard to age, body mass index (BMI), history of varicocele, his- tory of cryptorchidism, duration of infertility, results of genetic analysis, results of hormone profiles and testicu- lar histopathology results of the first mTESE. TESE technique On the day that the TESE operation was planned, addi- tional sperm samples were obtained and it was con- firmed that there were no sperms present. Informed con- sent was obtained from all of the patients before TESE. All of the patients underwent spinal anesthesia for TESE. A midline scrotal incision was made and the scrotal con- tent was pushed out from the side with the larger testis. The tunica vaginalis was opened and the tunica albug- inea that surrounds the testicle was visualized. After this stage, the operation was handled under operating micro- scope. As described by Schlegel, an avascular area was selected from the antimesenteric area to the tunica albug- inea and a 3 cm incision was made with a thin scalpel (4). Small samples were obtained from opaque, large, white tubules in the testicular parenchyma. Each sample was placed in a Petri dish filled with human tubal fluid. All samples were immediately evaluated by an embryol- ogist using a 200 x magnification microscope in order to investigate the presence of spermatozoa. The operation was terminated when suitable spermatozoa were found for ICSI. If spermatozoa were not detected in the first samples, additional samples were obtained from the same testicle. In cases where the spermatozoa were not found in the samples sent from the larger testis, the sam- ples were also obtained from the contra-lateral testis. The biopsy specimen was sent to the pathology laboratory intraoperatively in order to determine the testicular histopathology. Histopathological analysis In order to define the testicular histopathology, all testic- ular biopsy samples were fixed within Bouin's solution, and embedded into paraffin blocks following the tissue processing steps. 4 µm-thick sections were obtained, stained using hematoxylin and eosin dye, and evaluated under a microscope with 400 x magnification by the same pathologist who was experienced in this field for more than 10 years. Germinal epithelia of at least 100 seminif- erous tubules were evaluated for each biopsy sample. In the presence of germinal epithelium, the spermatoge- netic situation was assessed using the Johnsen's score (JS). According to JS, the tissue maturation and spermatoge- netic situation of the germinal epithelia of each sample were scored between 1 and 10. In this scoring system, tubular necrosis was scored as 1, Sertoli cell only was scored as 2, spermatogonia only was scored as 3, arrest at primary spermatocyte was scored as 4 or 5, arrest at the early spermatid stage was scored as 6 or 7, arrest at the late spermatid stage was scored as 8 or 9, and full sper- matogenesis was scored as 10 (5). The mean JS was cal- culated for each sample. Testicular biopsy specimens were classified according to the histopathological criteria as follows: normal spermatogenesis (NS) (mean JS; 10), hypospermatogenesis (HS) (mean JS; 8-9), late maturation arrest (LMA) (mean JS; 6-7), early maturation arrest (EMA) (mean JS; 3-4-5), Sertoli cell only (SCO) (mean JS; 2) and hyalinization of tubules (HT) (mean JS; 1). Statistical analysis The conformity of the variables to the normal distribu- tion was assessed with the Shapiro Wilk test. The cate- gorical variables were described using frequencies with percentages, and the numerical variables were described using the mean and standard deviation values. The Student’s t-test and the chi-square test were used for the intergroup analyses of the continuous variables. The chi-square test or the Fisher’s exact chi-square was used to for the categorical variables. More than two inde- pendent averages were compared with the ANOVA test and the Kruskal Wallis test. We performed the univari- ate and the multivariate analysis to identify the factors associated with and predictive of positive sperm retrieval during a salvage mTESE. Multiple logistic regression analysis was performed using a model including age, FSH and LH levels, and JS. The data analysis was carried out using the Statistical Package for the Social Science (SPSS Inc, Chicago, Illinois, USA) version 22.0 and a p value of < 0.05 was considered significant. RESULTS In 21 of the 49 participants (42.8%), sperm retrieval was possible via salvage mTESE. A statistically significant dif- ference was observed between the groups with and with- out sperm retrieval, with regard to FSH levels (20.4 ± 9.7 vs. 31.2 ± 10.4, respectively; p = 0.013) (Table 1). Yucel2_Stesura Seveso 28/06/18 16:40 Pagina 137 Archivio Italiano di Urologia e Andrologia 2018; 90, 2 C. Yücel, S. Budak, M. Zeynel Keskin, E. Kisa, Z. Kozacioglu 138 No significant difference was observed for the remaining hormone parameters, age, BMI, and duration of infertili- ty. The clinical and laboratory findings of the patients have been presented in Table 1. No history of cryp- torchidism was observed in any of the patients included in the study. Non-mosaic Klinefelter’s syndrome was observed in a patient prior to salvage mTESE, and no sperm could be obtained in this patient in salvage mTESE. Grade 1-2 varciocele was detected in the physical examination of 4 patients in the sperm retrieval group and 5 patients in the no sperm retrieval group. No serious complication was observed during mTESE or within the post-operative 3 weeks. The mean JS of the 49 participants was found to be 5.4 ± 1.1. The JS among the sperm retrieval group and the no sperm retrieval group were 4.2 ± 1.0 and 6.4 ± 1.1, respec- tively. The sperm retrieval rates in patients with HT, SCO, MA and HS histopathologies were 25% (2/8), 36% (4/11), 38% (8/21) and 60% (3/5), respectively. The sperm retrieval rate among patients with NS histopathology in the first mTESE was 100% (4/4). No statistically significant difference was observed between two groups with regard to testicular histopathology and mean JS (p = 0.621; p = 0.246, respec- tively). The multiple logistic regression analysis was performed by constructing a model including age, FSH and LH levels, and JS. This model has been presented in Table 2. It was observed that FSH was a significant and independent predictive factor for positive sperm retrieval in salvage mTESE (p = 0.032). DISCUSSION In this study, the factors predicting the success of positive sperm retrieval in salvage mTESE in patients with NOA and previous unsuccessful retrieval in mTESE were inves- tigated, and a statistical correlation was observed only between the preoperative FSH level and the success in sal- vage mTESE. Herein, we have summarized our experience of salvage mTESE in patients with NOA in our single unit within an 9-year interval. There are only five studies inves- tigating the factors predicting the success of salvage mTESE in patients with NOA and previous unsuccessful mTESE in the literature. In these studies, the sperm retrieval rates were reported to be between 30% and 46%. Table 1. Comparison between the successful and unsuccessful sperm retrieval in salvage mTESE. Variables Overall Spermatozoa No Spermatozoa P value (n=49) were retrievied were retrievied (n = 21) (n = 28) Age (years) 35.7 ± 5.1 35.4 ± 5.9 36.0 ± 4.9 0.817 Duration of infertility (years) 6.82 ± 3.67 6.58 ± 3.94 7.01 ± 3.63 0.844 BMI 22.8 ± 1.2 23.7 ± 1.3 22.1 ± 1.2 0.902 T (ng/dL) 401.1 ± 186.5 406.3 ± 279.0 397.3 ± 135.8 0.694 E2 (pg/mL) 31.3 ± 12.4 30.0 ± 9.9 32.4 ± 13.3 0.729 FSH (mIU/mL) 26.5 ± 10.2 20.4 ± 9.7 31.2 ± 10.4 0.013 LH (mIU/mL) 11.3 ± 7.1 8.7 ± 5.4 13.4 ± 9.1 0.161 PRL (ng/mL) 10.3 ± 3.8 10.5 ± 4.0 10.2 ± 3.9 0.834 Mean JS 5.4 ± 1.1 4.2 ± 1.0 6.4 ± 1.1 0.246 Histopathology (n/%) HT 8 (16.3) 2 (9.5) 6 (21.4) 0.621 SCO 11 (22.4) 4 (19.0) 7 (25) EMA 11 (22.4) 3 (14.2) 8 (28.5) LMA 10 (20.4) 5 (23.8) 5 (17.8) HS 5 (10.2) 3 (14.2) 2 (7.1) NS 4 (8.1) 4 (19.0) 0 (0) T, testosterone; E2, estradiol; FSH, follicle-stimulating hormone; LH, luteinizing hormone; PRL, prolactin; HT, hyalinization of tubules; SCO, Sertoli cell only; EMA, early maturation arrest; LMA, late maturation arrest; HS, hypospermatogenesis; NS, normal spermatogenesis; Table 2. Logistic regression analysis model for successful sperm retrieval in salvage mTESE. Variables OR 95% CI p value Age 0.992 0.938-1.136 0.657 FSH 0.963 0.944-0.982 0.032 LH 0.786 0.857-1.089 0.356 JS 0.763 0.865-0.979 0.812 JS, Johnsen's score; FSH, follicle-stimulating hormone; LH, luteinizing hormone; OR, odds ratio; CI, confidence interval. Table 3. Summary of the studies investigating the factors predicting the success of salvage TESE in patients diagnosed with NOA, who had previously undergone unsuccessful TESE. Variables Initial TESE Salvage TESE N SRR in SRR in the SRR in the SRR in the SRR in the SRR in the procedure procedure the study HT group SCO group MA group HS group NS group Okuba et al. (2002) Conventional Microscopic 13 30.7% - - - - - Tsujimara et al. (2006) Conventional Microscopic 46 45.7% - 39.1% 41.7% 100% - Ramasamy and Schlegel (2007) Conventional Microscopic 20 45% - 34.3% 61.55% 93.3% - Kalsi et al. (2015) Conventional Microscopic 58 46.55% - 40% 36.36% 75% - Xu et al. (2016) Conventional Microscopic 52 38.5% 25% 5.5% 25% 83.3% - This study Microscopic Microscopic 49 42.8% 25% 36% 38% 60% 100% HT, hyalinization of tubules; SCO, Sertoli cell only; EMA, early maturation arrest; LMA, late maturation arrest; HS, hypospermatogenesis; NS, normal spermatogenesis; SRR, sperm retrieval rate; TESE, testicular sperm extraction; N, number of patients included in the study. Yucel2_Stesura Seveso 28/06/18 16:40 Pagina 138 139Archivio Italiano di Urologia e Andrologia 2018; 90, 2 Predictive factors of successful salvage mTESE after failed mTESE Among those, the widest study included 58 patients. The outcomes of these studies have been summarized in Table 3 (6-10). In all of these studies, the first unsuccess- ful TESE was performed via the conventional procedure, whereas salvage TESE was performed with the guidance of a microscope. Therefore, our study is the first to investi- gate the factors affecting the success of salvage mTESE in patients with NOA and previous unsuccessful TESE per- formed using the microscopic procedure. Although empirical medical treatments may be given in patients with NOA prior to mTESE in order to increase the success rates, the efficacies of these treatments have not been confirmed via randomized controlled studies (11). Some authors claim that the possibility to retrieve sperm in salvage TESE may be increased via hormone therapies using clomiphene or human chorionic gonadotropin (HCG) following unsuccessful TESE (12). However, due to the low level of evidence and lack of recommendations for these therapies in the guidelines, no hormone therapy was performed in the time between the first unsuccessful and salvage mTESEs. There are conflicting data about the effect of FSH level on the success of mTESE in the literature. There are studies demonstrating no effect of FSH on the success of mTESE, whereas there are others demonstrating the con- trary (13). FSH acts by binding to its receptors on the Sertoli cells, which are important for spermatogenesis in the testis. Therefore, it tends to decrease in patients with impaired spermatogenesis. Although high FSH levels were related to global impairment of spermatogenesis, there may be normal foci of spermatogenesis in the testi- cles of these patients (14). Xu et al. and Kalsi et al. have reported no significant difference between the FSH lev- els of the patients with or without sperm retrieval in sal- vage TESE (9, 10). On the contrary, in our study, the FSH levels in patients with no sperm retrieval in salvage TESE was observed to be higher. TESE is an invasive procedure that may lead to compli- cations such as hematoma, infection, fibrosis and even permanent devascularisation. Diagnostic testicular biop- sy has complications similar to the mTESE operation. Furthermore, sperm retrieval in subsequent mTESE can- not be assured by retrieved sperm in diagnostic testicu- lar biopsy in patients with NOA. Diagnostic testicular biopsy has not been recommended in clinical practice due to the additional cost, repetitive surgical procedures and the invasive nature of the procedure that increase the risk of complications (15). Thus, the diagnostic tes- ticular biopsy procedure is not being performed in our clinics prior to TESE. Testicular biopsy samples have been obtained during the initial TESE surgery. Additionally, one of the strong aspects of our study was that the testicular histopathology was evaluated by the same and experienced pathologist.Tsujimara et al., Ramasay et al., and Kalsi et al. have evaluated testicular histopathology by classifying into SCO, MA and HS sub- groups in order to assess its predictive value for the out- come of salvage TESE (7-9). In addition to these three studies, Xu et al. have evaluated HT testicular histopathology as an individual subgroup (10). In contrast to these studies, the NS testicular histopathol- ogy subgroup was individually evaluated in our study, and patients with MA were divided into the LMA and EMA groups. This classification was made using the JS in our study and therefore, testicular histopathology is believed to be subgrouped more accurately. Tsujimara et al., Ramasay et al. and Kalsi et al. compared testicular histopathologies, and determined the sperm retrieval rates of 39.1%, 34.3% and 40%, respectively, in patients with SCO histopatholo- gy prior to salvage TESE (7-9). In our study, the sperm retrieval rate was 36% in patients with SCO histopatholo- gy. In the study of Xu et al., sperm retrieval was possible in 25% of the patients with HT histopathology in salvage TESE (10). Similiarly, in our study, the sperm retrieval rate among patients with HT histopathology was 25%. These results demonstrate that the possibility of sperm retrieval continues in salvage mTESE even after a previous unsuc- cessful TESE. In our study, the sperm retrieval rate in sal- vage mTESE in patients with NS histopathology was 100% and it was 60% in patients with HS histopathology. Compared to the other studies in the literature, the sperm retrieval rate among patients in the HS subgroup of our study was lower. The reason for this difference may be the different subgrouping in our study to that in the literature. It was also concluded that the testicular histopathology was not a predictive factor for the success of salvage TESE. The sample size in our study and those in other studies in the literature were small, it is believed that accurate results may be accessed via meta-analyses performed in the future. Our study has some limitations. First, it was a retrospec- tive study. The effects of cryptorchidism and Klinefelter’s syndrome could not be evaluated in salvage TESE, since there was no history of cryptorchidism and only one patient had a history of Klinefelter’s syndrome in our study. Although short-term complications following sal- vage TESE were evaluated, no long-term evaluation was carried out, which is the second limitation of our study. Not all sperms retrieved in TESE can be used in the ICSI procedure. Reproductive analysis of the patients follow- ing ICSI was not included in our study, which may be considered as another limitation; however, the outcomes of ICSI are affected by many factors including those of the women as well, and since the priority of our study was the factors affecting the success in salvage mTESE, the ICSI results were not included in the study. CONCLUSIONS Our study suggests that salvage mTESE is a safe alterna- tive treatment method in patients with NOA and previ- ous unsuccessful TESE, since the sperm retrieval rate was relatively higher. Evaluation of the preoperative FSH levels may be useful in determining the best candidates for this patient group. Further multi-center, prospective studies with larger sample sizes should be conducted in order to better understand the subject. ACKNOWLEDGEMENTS We would like to thank to Mustafa Karabicak, Ertan Can, Ozgur Cakmak for their contribution to the statistical analysis and Can Kose, Ulku Kucuk, Gokhan Koc for their contribution to the acquisition of data. Yucel2_Stesura Seveso 28/06/18 16:40 Pagina 139 Archivio Italiano di Urologia e Andrologia 2018; 90, 2 C. Yücel, S. Budak, M. Zeynel Keskin, E. Kisa, Z. Kozacioglu 140 REFERENCES 1. Keskin MZ, Budak S, Aksoy EE, et al. Investigation of the effect of body mass index (BMI) on semen parameters and male reproductive system hormones. Arch Ital Urol Androl. 2017; 89:219-21. 2. Yucel C, Keskin MZ, Cakmak O, et al. Predictive value of pre- operative inflammation-based prognostic scores (neutrophil-to-lym- phocyte ratio, platelet-to-lymphocyte ratio, and monocyte-to- eosinophil ratio) in testicular sperm extraction: a pilot study. Andrology. 2017; 5:1100-4. 3. Vloeberghs V, Verheyen G, Haentjens P, et al. How successful is TESE-ICSI in couples with non-obstructive azoospermia? Hum Reprod. 2015; 30:1790-1796. 4. Schlegel PN. Testicular sperm extraction: microdissection improves sperm yield with minimal tissue excision. Hum Reprod. 1999; 14:131-5. 5. Ustuner M, Yılmaz H, Yavuz U, et al. Varicocele repair improves testicular histology in men with nonobstructive azoospermia. Biomed Res Int. 2015; 709452. 6. Okada H, Dobashi M, Yamazaki T, et al. Conventional versus microdissection testicular sperm extraction for nonobstructive azoospermia. J Urol. 2002; 168:1063-7. 7. Tsujimura A, Miyagawa Y, Takao T, et al. Salvage microdissec- tion testicular sperm extraction after failed conventional testicular sperm extraction in patients with nonobstructive azoospermia. J Urol. 2006; 175:1446-9. 8. Ramasamy R, Schlegel PN. Microdissection testicular sperm extraction: effect of prior biopsy on success of sperm retrieval. J Urol. 2007; 177:1447-9. 9. Kalsi JS, Shah P, Thum Y, et al. Salvage micro-dissection testicu- lar sperm extraction; outcome in men with non-obstructive azoospermia with previous failed sperm retrievals. BJU Int. 2015; 116:460-5. 10. Xu T, Peng L, Lin X, et al. Predictors for successful sperm retrieval of salvage microdissection testicular sperm extraction (TESE) following failed TESE in nonobstructive azoospermia patients. Andrologia. 2017; 49: e12642. 11. Patel DP, Chandrapal JC, Hotaling JM. Hormone-based treat- ments in subfertile males. Curr Urol Rep. 2016; 17:1-8. 12. Hussein A, Ozgok Y, Ross L, et al. Optimization of spermatoge- nesis-regulating hormones in patients with non-obstructive azoospermia and its impact on sperm retrieval: a multicentre study. BJU Int. 2013; 111:110-4. 13. Ramasamy R, Lin K, Gosden LV, et al. High serum FSH levels in men with nonobstructive azoospermia does not affect success of microdissection testicular sperm extraction. Fertil Steril. 2009; 92:590-3. 14. Hung AJ, King P, Schlegel PN. Uniform testicular maturation arrest: a unique subset of men with nonobstructive azoospermia. J Urol. 2007; 178: 608-12. 15. Haimov-Kochman R, Lossos F, Prus D, et al. The value of repeat testicular sperm retrieval in azoospermic men. Fertil Steril. 2009; 91:1401-3. Correspondence Cem Yücel, MD (Corresponding Author) meclecuy@hotmail.com Salih Budak, MD salihbudak1977@gmail.com Mehmet Zeynel Keskin, MD zeynel_akd@hotmail.com Erdem Kisa, MD drerdemkisa@hotmail.com Zafer Kozacıoglu, MD Associate Prof. zaferkozacioglu@gmail.com Tepecik Training and Research Hospital Yenisehir Mah, Gaziler Cad. No:468, Konak/Izmir, Turkey Yucel2_Stesura Seveso 28/06/18 16:40 Pagina 140