A retrospective cohort study was performed at the Shahid Beheshti Hospital in Isfahan, Iran. A total of 170 patients with NOA who were referred for fertility treatment from 2013 to 2016 were included in this study. Informed consent was obtained from all patients before inclusion in the study. All patients were confirmed to have NOA using at least two semen analyses according to World Health Organization guidelines [14], without any sperm and with normal pH and volume (>1.5 mL), indicating both functional seminal vesicles and open ejaculatory ducts. Additionally, the assessment of diagnostic parameters, including clinical history, physical examination, hormonal analysis, and pathology assay should be performed to exclude obstructive azoospermia.

For the hormonal assay, serum FSH, LH, testosterone, and prolactin concentrations were measured by an electrochemical luminescence assay in the morning. The reference ranges were considered to be 1.5 to 12.45 mIU/mL for FSH, 1.7 to 8.6 mIU/mL for LH, 2.84 to 8 ng/mL for testosterone, and 1.8 to 20.3 ng/mL for prolactin. Testicular volume was measured by physical examination using an orchidometer. A volume of <15 mL per testis was considered to indicate small testicular size. The mean body mass index of the patients was 25.2±3.9 kg/m² (range, 16–42 kg/m²) and the couples had a median duration of infertility of 3.5 years.

The TESE and micro-TESE procedures were performed according to the standard protocols described by Schlegel for micro-TESE [15]. Briefly, for both surgical procedures, a median raphe incision was made in the scrotum, the tunica vaginalis was opened, and the testis was delivered. In TESE procedures, a longitudinal testicular biopsy was obtained according to the size of the testis (at least 10 mm×5 mm). In micro-TESE procedures, using an operating microscope with ×25 to ×40 magnification, enlarged and opaque seminiferous tubules were removed. Then, the sperm recovery procedure was carried out based on the standard protocol by the embryologist. If any viable sperm were present in a wet preparation, they were used or cryopreserved for future ICSI treatment. Furthermore, the histopathology specimen was analyzed. Based on the most predominant histopathological patterns, patients with NOA were classified according to testicular histology as being affected by hypospermatogenesis, MA, or SCOS. In this study, the MA group is defined by either complete or partial MA. The hypospermatogenesis group includes mild to severe hypospermatogenesis. The presence of one or more spermatozoon in the testicular extraction was considered as successful sperm retrieval or a successful TESE outcome.

Karyotyping was performed according to standard methods on phytohemagglutinin-stimulated peripheral lymphocyte cultures [16]. Blood cells were cultured at Roswell Park Memorial Institute 1640 medium, including 10% fetal bovine serum and 2% phytohemagglutinin at 37℃. After 68 to 72 hours, Colcemid (Sigma-Aldrich, St. Louis, MO, USA) was added to the cell. Then, the cultures were harvested and cells in metaphase were stained with trypsin G banding. Chromosomal abnormalities were designated using the standard International System for Human Cytogenetic Nomenclature guidelines [17].

Summary measures for the factors, according to sperm retrieval and pathology, are presented in Tables 1 and 2, respectively. Binary and multinomial logistic regression models were evaluated for sperm retrieval and pathology as the response variables, respectively. For each model, independent variables and the estimated coefficients, odds ratios (ORs), and corresponding confidence intervals are given in Tables 2 and 3.

All analyses were carried out with SPSS ver. 16.0 (SPSS Inc., Chicago, IL, USA) and the significance level was considered to be p<0.05.

The mean age of all patients was 32.56±5.53 years. Based on the histopathological findings, the 170 patients were classified into three groups: 74 (43.5%) cases of SCOS, 46 (27.1%) cases of MA, and 50 (29.4%) cases of hypospermatogenesis. The levels of FSH were significantly in excess of the normal ranges in the SCOS and MA groups when compared to the hypospermatogenesis group (MA: OR, 1.146, p=0.034; SCOS: OR, 1.190, p=0.007). In addition, prolactin levels were significantly higher in the SCOS group than in the hypospermatogenesis group (OR, 1.236; p=0.006), as shown in Table 2. No associations were observed among age, chromosomal abnormalities, and histopathological patterns (p>0.05) (Table 2). The rate of chromosomal abnormalities was 12.9% in all patients. The greatest number of patients with chromosomal abnormalities (17 cases) showed 47,XXY karyotyping. Other chromosomal abnormalities included 47,XYY (two cases), 46,XX male (one case), 46X,+mar (one case) and 46,XY,DelY q11.2 (one case).

Our results showed that the overall mean rate of sperm retrieval was 48.8%. These rates were 21.6% in the SCOS group, 43.5% in the MA group, and 94% in the hypospermatogenesis group. Based on the multinomial logistic regression model, the rates of unsuccessful sperm retrieval were significantly higher in the MA and SCOS groups than in the hypospermatogenesis group (Table 2). The logistic regression model confirmed that high levels of FSH and small testicular volume were significantly associated with lower chances of successful sperm retrieval (OR, 0.717, p<0.001; OR, 10.5, p<0.01, respectively) (Table 3). However, age, testosterone, and prolactin levels were not associated with positive sperm retrieval (p>0.05). The results exhibited statistically significant differences in sperm retrieval between conventional TESE and micro-TESE procedures, and there was an eight-fold greater chance of sperm retrieval with the micro-TESE method than with conventional TESE. The association between chromosomal abnormalities and the rate of sperm retrieval was assessed. No associations were observed among the aforementioned factors and sperm retrieval (p>0.05).