Azoospermia is defined based on the absence of spermatozoa in a given ejaculate. Proper laboratory technique is crucial to reduce analytical error and enhance precision when analyzing semen specimens (Aziz 2013; Esteves et al. 2012). Ejaculates of men with spermatogenic failure usually have normal volume and pH, which indicates both functional seminal vesicles and patent ejaculatory ducts. The lower reference limits for ejaculate volume and pH are 1.5 ml (fifth percentile, 95 % confidence interval 1.4–1.7) and 7.2, respectively (Cooper et al. 2010). Retrograde ejaculation should be suspected when a given ejaculate volume is <1 ml, and the diagnosis is confirmed by the finding of spermatozoa in the post-ejaculatory urine (Esteves et al. 2011a).

The assessment of an initially normal-volume azoospermic ejaculate should be immediately followed by the examination of the pelleted semen to exclude cryptozoospermia, which is defined by the presence of sperm only in the centrifuged pellet (Aziz 2013). In one study, centrifuging semen at the low speed of 200 g for 10 min revealed that 22.8 % of men diagnosed with azoospermia had spermatozoa in the semen pellet (Jaffe et al 1998). In addition, when supernatants resulting from low-speed centrifugation were centrifuged at higher speeds (>1000 g) for longer periods, spermatozoa were also detected (Corea et al. 2005). Thus, the accuracy of any centrifugation protocol of less than 1000 g in pelleting all the spermatozoa in an ejaculate is uncertain (Cooper et al. 2006). The importance of finding such minimal number of sperm is to allow assisted reproductive techniques (ART) to be performed with ejaculated sperm, thus avoiding the more invasive sperm retrieval methods. At our institution, we perform centrifugation at 3000 g for 15 min, which is followed by a careful examination of the pellet for the presence of sperm. Moreover, the diagnosis of azoospermia should be based on the examination of multiple semen specimens as transient azoospermia secondary to toxic, environmental, infectious, or iatrogenic conditions may occur (Castilla et al. 2006; Keel 2006). The examination of ejaculates on multiple occasions is also important given the large biological variability in semen specimens from the same individuals (Esteves et al. 2012; Castilla et al. 2006; Keel 2006). Patients should receive clear instructions on how to collect the entire ejaculate and to report the loss of any fraction of the sample. Determination of fructose, a major component of seminal vesicle secretion, is usually not necessary because the presence of a normal-volume ejaculate coupled with normal pH practically excludes any problem at the ejaculatory ducts or seminal vesicle level (Esteves et al. 2011a). In summary, azoospermia should be defined based on the absence of spermatozoa on multiple semen examinations after centrifugation of complete semen specimens using microscopic analysis.

History and physical examination and hormonal analysis (follicle-stimulating hormone and total testosterone serum levels) are undertaken to define the type of azoospermia. Together, these factors provide a >90 % prediction of the type of azoospermia (obstructive vs. nonobstructive). Obstructive azoospermia (OA) is attributed to a mechanical blockage that can occur anywhere along the reproductive tract, including the vas deferens, epididymis, and ejaculatory duct. OA is considered to be one of the most favorable prognostic conditions in male infertility since spermatogenesis is not disrupted, unlike spermatogenic failure (Esteves et al. 2013a; American Society for Reproductive Medicine and Society for Male Reproduction and Urology 2008). Etiology conditions associated with nonobstructive azoospermia (NOA) include genetic and congenital abnormalities, postinfections, exposure to gonadotoxins, medications, varicocele, trauma, endocrine disorders, and idiopathy. A detailed medical history should be obtained for any factor that may cause spermatogenic failure. Information not exclusive of the following areas should be collected: (a) previous diseases during childhood and puberty such as viral orchitis and cryptorchidism; (b) surgeries performed, especially those involving the pelvic and inguinal regions and genitalia; (c) genital traumas; (d) infections such as epididymo-orchitis and urethritis; (e) physical and sexual development; and (f) exposure to gonadotoxic agents such as radiotherapy, chemotherapy, and steroid abuse (Esteves et al. 2011a; Carpi et al. 2009).

Physical examination in men with spermatogenic failure usually reveals normal epididymides and palpable vasa deferentia. Small-sized testes (<15 ml in volume) are often encountered as approximately 85 % of the testicular parenchyma is involved in spermatogenesis. Nevertheless, testicular size is not a reliable clinical marker of sperm production. Men with spermatogenic maturation arrest, in whom spermatogenesis is hampered prior to its completion, have well-developed and normal-volume testes (Sokol and Swerdloff 1997; Hung et al. 2007).

The serum levels of follicle-stimulating hormone (FSH) are usually elevated, while testosterone is either low (<300 ng/dl) or within lower limits in men with spermatogenic failure (Gudeloglu and Parekattil 2013; Esteves et al. 2011a). FSH levels greater than twice the upper normal limit are a reliable indicator of spermatogenic failure (American Society for Reproductive Medicine and Society for Male Reproduction and Urology 2008). In one report, low testosterone levels were found in 45 % of the males with NOA who visited an infertility clinic (Sussman et al. 2008). In another study evaluating hormonal data of 736 men with NOA who were candidates for sperm retrieval, 346 (47 %) had baseline total testosterone (TT) levels <300 ng/dL (Reifsnyder et al. 2012). Low testosterone levels often reflect Leydig cell insufficiency, which is accompanied by elevated (or within upper limits) luteinizing hormone (LH) levels (Bobjer et al. 2012; Reifsnyder et al. 2012). Nevertheless, low testosterone levels in men with SF may also result from obesity and metabolic dysfunction (Kumar 2013). Obesity is associated with an increased serum estradiol levels due to the increased peripheral aromatization of C19 androgens (androstenedione, T) under the influence of aromatase, a product of the CYP19 gene, especially in individuals with tetranucleotide TTTA repeat polymorphism (TTTAn) present in intron 4 of the CYP19 gene (Hammoud et al. 2010). Elevated estradiol levels suppress pituitary LH and FSH secretion and also directly inhibit testosterone biosynthesis (Kumar 2013; Tchernof et al. 1995). Furthermore, Isidori et al. (1999) have demonstrated that excess circulating leptin may be an important contributor to the reduced androgen serum levels in male obesity. Their data have indicated that leptin has negative actions on steroidogenesis that are mediated by specific receptors in the Leydig cells. Low testosterone levels could also reflect an adaptation to changed SHBG levels and not testosterone deficiency. In fact, Strain et al. (1994) have reported that, during weight loss, serum SHBG levels increase at an average slope of 0.43 nmol/L per unit decrease in body mass index (BMI). Hence, the increased serum TT concentrations as seen after weight loss may be due to a combination of mechanisms that include (i) an increased binding capacity of SHBG, (ii) an increased amplitude of spontaneous LH pulses, (ii) a decreased androgen aromatization, and (iv) a decrease in circulating leptin and insulin concentrations. Surprisingly, a normal endocrine profile can be also found in men with spermatogenic failure. Control feedback of FSH and LH secretions is based on the number of spermatogonia and Leydig cells, respectively, which is well preserved in men with maturation arrest. It has been reported that patients with diffuse spermatogenic maturation arrest and 10 % of those diagnosed with Sertoli-cell-only syndrome (SCOS) present with nonelevated endogenous gonadotropins (Sokol and Swerdloff 1997; Hung et al. 2007).

Lastly, it is important to differentiate azoospermia due to spermatogenic failure from azoospermia due to hypogonadotropic hypogonadism (HH) as both conditions fall in the category of nonobstructive azoospermia (NOA). HH is an endocrine disorder characterized by failure of spermatogenesis due to lack of appropriate stimulation by gonadotropins, while spermatogenic failure comprises the most severe conditions associated with an intrinsic testicular impairment (Fraietta et al. 2013). Men with NOA due to HH have remarkably low levels of pituitary gonadotropins (FSH and LH levels below 1.2 mUI/ml) and androgens and usually have signs of absent or poor virilization. This category of NOA includes not only patients with congenital forms of HH but also men whose spermatogenic potential has been suppressed by excess exogenous androgen administration. Although it is out of my scope to discuss HH in detail, it is worth to mention that patients with HH, albeit rarely seen in the clinical settings, benefit from specific hormonal therapy and often show remarkable recovery of spermatogenic function with exogenously administered gonadotropins or gonadotropin-releasing hormone (Fraietta et al. 2013).

The “gold standard” test for confirmation of azoospermia due to SF is testicular biopsy and histopathology analysis. Hypospermatogenesis, germ cell maturation arrest, germ cell aplasia (Sertoli-cell-only syndrome), tubular sclerosis, or a combination of those is usually found on the histological examination of testicular biopsy specimens in spermatogenic failure. Biopsies can be performed using percutaneous or open methods. Histopathology results have been used not only to confirm the diagnosis of SF but also to predict the chances of finding testicular sperm on retrievals. In a recent study from our group evaluating 356 men with spermatogenic failure, patients with Sertoli-cell-only had lower sperm retrieval rates (19.5 %) compared with those with maturation arrest (40.3 %, P = .007), and both categories had lower sperm retrieval rates (SRR) compared with hypospermatogenesis (100.0 %, P < .001; Esteves and Agarwal 2014). Although our data indicate that histopathology phenotypes have prognostic value, caution should be applied when interpreting results because an advanced site of sperm production can be found even in SCO, which represents the worst histopathology phenotype, in approximately 20 % of the cases (Esteves et al. 2014; Esteves and Agarwal 2014; Ashraf et al. 2013; Verza Jr and Esteves 2011). Removal of testicular tissue with the sole purpose of histopathology evaluation could potentially remove the rare foci of sperm production and thus jeopardize the chances of future retrieval attempts (Esteves et al. 2011b). Hence, we do not recommend routine testicular biopsy prior to sperm retrieval. We only perform testicular biopsies when a differential diagnosis between obstructive and nonobstructive azoospermia could not be established. In these cases, our approach is to perform the procedure either using a percutaneous or an open-“window” technique without testis delivery (Esteves et al. 2011a; Esteves and Verza 2012). Specimens should be placed in a fixative solution such as Bouin’s, Zenker’s, or glutaraldehyde; formalin should not be used as it may disrupt the tissue architecture. A fragment is taken for wet examination in addition to conventional histopathology analysis. When mature sperm is found on a wet examination, we routinely cryopreserve testicular spermatozoa using the liquid nitrogen vapor technique (Esteves and Verza 2012; Esteves and Varghese 2012).

In conclusion, proper laboratory techniques are needed to reduce the amount of analytical error and enhance sperm count precision when evaluating azoospermic specimens. The correct assessment of an initially azoospermic semen should be followed by the examination of multiple specimens after centrifugation to exclude cryptozoospermia, which is defined by presence of a very small number of live sperm in a centrifuged pellet. Accurate assessment of very low sperm counts is aimed to avoid labeling men with very low sperm counts as azoospermic, and it is particularly important in the current era of ART. History and physical examination and hormonal analysis are undertaken to define the type of azoospermia, which provide high diagnostic accuracy to discriminate azoospermia due to spermatogenic failure from obstructive azoospermia and hypogonadotropic hypogonadism (Table 7.1). Although the “gold standard” diagnostic test in azoospermia related to spermatogenic failure is testicular biopsy, removal of testicular tissue with the sole purpose of histopathology evaluation could potentially remove the rare foci of sperm production and thus jeopardize the chances of future retrieval attempts. Testicular biopsy prior to sperm retrieval is therefore not routinely recommended. Testicular biopsy can be performed in selected cases provided a wet prep examination and sperm cryopreservation is available.

Owed to the untreatable nature of spermatogenic failure, sperm retrieval (SR) and ART are the only options for these men to generate their own biological offspring. Uncertainty of sperm acquisition, however, makes prognostic factors very desirable. Though factors such as etiology, testicular volume, serum levels of pituitary gonadotropins, and testicular histopathology results reflect a global spermatogenic function, they cannot accurately discriminate individuals in whom foci of sperm production will be found upon SR. In an early series involving 60 men with SF, we determined the accuracy of commonly used prognostic parameters using a logistic regression analysis (Verza Jr and Esteves 2011). We confirmed that these parameters have low accuracy as the areas under the receiver-operating characteristic (ROC) curves of FSH, testosterone, and testicular volume for predicting a positive sperm extraction were 0.53, 0.59, and 0.52, respectively. In another study, Tournaye and cols. combined clinical and laboratory parameters, such as testicular volume and FSH levels and histopathology results, and found that diagnostic accuracy was only 74 % (Tournaye et al. 1997). Testicular sperm have been obtained in different etiology categories, including cryptorchidism, post-orchitis, Klinefelter syndrome, radio-/chemotherapy, and idiopathy, with variable success rates ranging from 25 to 70 % (Esteves et al. 2010; Schiff et al. 2005; Chan et al. 2001; Raman and Schlegel 2003; Esteves 2013). In summary, clinical parameters and endocrine profile are unreliable markers for determining the chances of sperm acquisition in men with azoospermia due to spermatogenic failure.

In contrast, the molecular diagnosis and subtyping of Y chromosome microdeletions (YCMD) have been shown to be useful preoperative biomarkers to determine the chances of sperm retrieval in men with azoospermia due to YCMD (Esteves and Agarwal 2011; Stahl et al. 2010; Krausz et al. 2000; Peterlin et al. 2002; Hopps et al. 2003; Simoni et al. 2008; Kleiman et al. 2011, 2012; Hamada et al. 2013). A microdeletion is a chromosomal deletion that usually spans over several genes but is small in size and cannot be detected using conventional cytogenetic methods such as karyotyping (Navarro-Costa et al. 2010; Hamada et al. 2013). The long arm of the Y chromosome contains a region at Yq11 that clusters 26 genes involved in spermatogenesis regulation (Simoni et al. 2008; Hamada et al. 2013; Repping et al. 2002; Krausz et al. 2014). This region is referred to as “azoospermia factor” (AZF) because microdeletions at this interval are often associated with azoospermia (Fig. 7.2). The application of molecular technology has allowed the recognition of three AZF subregions designated as AZFa, AZFb, and AZFc, each one including a major AZF candidate gene (Simoni et al. 2008; Krausz et al. 2014). It has been estimated that approximately 10 % of men with azoospermia due to spermatogenic failure harbor microdeletions within the AZF region that might explain their condition (Simoni et al. 2008; Krausz et al. 2014).

From the medical point of view, the following microdeletions have recurrently been found in men with spermatogenic failure (Krausz et al. 2014; Navarro-Costa et al. 2010): (i) AZFa, (ii) AZFb (P5/proximal P1), (iii) AZFbc (P5/distal P1 or P4/distal P1), and (iv) AZFc (b2/b4). The most frequent deletion subtypes comprise the AZFc region (~80 %) followed by AZFa (0.5–4 %), AZFb (1–5 %), and AZFbc (1–3 %) regions (Krausz et al. 2014). Deletions differentially affecting these AZF subregions cause a distinct disruption of germ cell development. AZFa deletions that remove the entire AZFa are invariably associated with the testicular histopathology phenotype of pure SCOS with no residual areas of active spermatogenesis. Although partial AZFa deletions have been described and may be eventually associated with residual spermatogenesis, this event is extremely rare (Tyler-Smith and Krausz 2009). Hence, the diagnosis of a deletion in the AZFa region implies that the chances of retrieving testicular spermatozoa for ICSI are virtually nonexistent (Krausz et al. 2000; Hopps et al. 2003; Simoni et al. 2008; Kleiman et al. 2011; Vogt and Bender 2013). The clinical feature of complete AZFb and AZFbc (P5/proximal P1, P5/distal P1, P4/distal P1) deletions is similar to AZFa deletions as the chances of finding spermatozoa on attempts of sperm retrieval are close to zero (Krausz et al. 2000; Hopps et al. 2003; Kleiman et al. 2011). In AZFb and AZFbc deletions, the most common testicular histopathology phenotype is spermatogenic maturation arrest, but SCOS can also be found. Nevertheless, spermatid arrest and crypto-/oligozoospermia have been reported in three patients with a complete AZFb or AZbc deletions (Soares et al. 2012; Longepied et al. 2010). In addition, spermatozoa have been identified in rare cases of complete and partial AZFb and AZFbc deletions (Kleiman et al. 2011). At present, however, given the difficulties to explain the biological nature of these unusual phenotypes, it is sound to assume that the diagnosis of complete deletions of AZFb or AZFbc (P5/proximal P1, P5/distal P1, P4/distal P1) implies that the chances of a successful testicular sperm retrieval are virtually zero (Krausz et al. 2014). In contrast, the chances of successful sperm retrieval in men with NOA and AZFc deletions are 50–70 % (Peterlin et al. 2002; Simoni et al. 2008). AZFc deletions are usually associated with residual spermatogenesis, and therefore testicular spermatozoa can be surgically retrieved and children can be conceived by ICSI (Kent-First et al. 1996; Mulhall et al. 1997; Kamischke et al. 1999; van Golde et al. 2001; Oates et al. 2002). The probability of fatherhood by ICSI seems to be unaltered by the presence of AZFc microdeletions (Peterlin et al. 2002; Kent-First et al. 1996; Mulhall et al. 1997; Kamischke et al. 1999; Oates et al. 2002; Cram et al. 2000). Notwithstanding, some authors have reported impaired embryo development in such cases (Simoni et al. 2008; van Golde et al. 2001). The male offspring born via ICSI from fathers with AZFc microdeletions will inherit the Yq microdeletion and as a result infertility. However, the exact testicular phenotype cannot be predicted as AZFc deletions may jeopardize Y chromosome integrity, predisposing to chromosome loss and sex reversal. There is a potential risk for the 45,X0 karyotype and to the mosaic phenotype 45,X/46,XY in these offspring, which may lead to spontaneous abortion or a newborn with genital ambiguity (Siffroi et al. 2000; Patsalis et al. 2000; Rajpert-De Meyts et al. 2011). Genetic counseling is therefore mandatory to provide information about the risk of conceiving a son with infertility and other genetic abnormalities.

Diagnostic testing for YCMD is based on a multiplex polymerase chain reaction (PCR) blood test aimed to amplify the AZFa, AZFb, and AZFc regions of the Y chromosome (Hamada et al. 2013). This technique primarily amplifies anonymous sequences of the Y chromosome using specific sequence-tagged sites (STSs) primers that are not polymorphic and are well known to be deleted in men affected by azoospermia according to the known, clinically relevant microdeletion pattern (Krausz et al. 2014). To obtain uniform results, it is necessary to follow validated guidelines, such as those issued by the European Association of Andrology (EAA) and the European Molecular Genetics Quality Network (EMQN) (Krausz et al. 2014). The basic set of PCR primers recommended by the EAA/EMQN to be used in multiplex PCR reactions for the diagnosis of Yq microdeletion includes sY14 (SRY), ZFX/ZFY, sY84 and sY86 (AZFa), sY127 and sY134 (AZFb), and sY254 and sY255 (AZFc) (Fig. 7.2). While the primer for the SRY gene is included as a control for the testis-determining factor on the short arm of the Y chromosome, the primers for the ZFX/ZFY gene act as internal controls because these primers amplify a unique fragment both in male and female DNA, respectively. A DNA sample from a fertile male and from a woman and a blank (water) control should be run in parallel with the set of primers. According to the current knowledge, once a deletion of both primers within a region is detected, the probability of a complete deletion is very high. The use of the aforementioned primer set enables the detection of almost all clinically relevant deletions and of over 95 % of the deletions reported in the literature in the three AZF regions (Krausz et al. 2014). However, as partial AZFa, AZFb, and AZFbc deletions have been described and their phenotypic expression is milder than the complete ones (Krausz et al. 2000; Kleiman et al. 2011), the definition of the extension of the deletion is now recommended in sperm retrieval candidates and should be based on additional markers as described by Krausz and colleagues (2014).

In conclusion, patients with azoospermia due to spermatogenic failure who are candidates for sperm retrieval and ICSI should be screened for Y chromosome microdeletions because the diagnosis of a deletion has prognostic value and influences therapeutic options (Table 7.1). Retrieval attempts are not recommended in cases of complete deletion of the AZFa region. Sperm retrieval in azoospermic carriers of deletions of the AZFb or AZFbc regions may be eventually attempted. However, the patient should be fully informed about the very low/virtually zero chance to retrieve spermatozoa. Owed to reports of deletion carriers among men with nonidiopathic NOA, including cryptorchidism, post-chemo-/radiotherapy, varicocele, and Klinefelter syndrome, the presence of any of these diagnosis categories accompanied by azoospermia should be an indication for YCMD screening testing (Krausz et al. 1999; Mitra et al. 2006). Genetic counseling should be offered to men with AZFc deletions who are candidates for sperm retrieval because testicular spermatozoa used for ICSI will invariably transmit the deletion from father to son. Although the likely result is azoospermia, AZFc microdeletions might be associated with an increased risk of miscarriage and other genetic abnormalities in the offspring.