Implementing advanced assisted reproductive technologies (ARTs) significantly increased the incidence of pregnancy in married fertility-challenged couples with azoospermia in male partners. It is important to differentiate obstructive azoospermia (OA) and non-obstructive azoospermia (NOA) because the difference between them is critical. OA is a relatively benign condition with excellent reproductive outcomes. Sometimes, it can be successfully reconstructed surgically [1-4]. On the other hand, NOA is complicated in treatment, and in most cases, it requires technically advanced Microdissection Testicular Sperm Extraction (microTESE) with further Intracytoplasmic Sperm Injection (ICSI), which provides relatively moderate outcomes [5][6]. Thus, the difference between OA and NOA plays an important role in patient counseling to provide adequate expectations.

NOA is observed more often than OA. Its etiology may remain unclear despite a wide diagnostic search. On the contrary, in most cases, OA has a clearer etiology because of its congenital or acquired nature. Visualizing the study methods can reveal the precise area of obstruction, which can affect the treatment tactics. Differential diagnosis between OA and NOA is based on the study of medical history, the results of the physical and instrumental examination, spermogram, the level of serum hormones, and the results of genetic screening [6-11]. It was shown that 96.0% of patients with OA had the levels of follicle-stimulating hormone (FSH) < 7.6 IU/l and a normal volume of testicles (long axis length > 4.6 cm), while 89.0% of patients with NOA had an elevated level of FSH and a lower volume of testicles [12]. This rule is used in clinical practice for differential diagnostics between OA and NOA. Still, there is some uncertainty.

Sometimes, the difference between OA and NOA is less clear due to long-term obstruction of the seminiferous ducts [13][14]. Furthermore, some cases of NOA with a normal volume of testicles and a serum hormone level that varies from normal to threshold can mimic OA [15]. A urologist-andrologist can face complicated cases with unexpectedly low sperm count (or even lack of sperm) during surgical sperm retrieval attempts in patients with suspected OA. In such cases, in patients with testicle pathology, changes can be observed that are typical of spermatogenic insufficiency.

The study aimed to describe, analyze, and classify potential issues in differential diagnosis of two types of azoospermia.

The authors performed a retrospective analysis of the database that included patients who underwent surgical sperm retrieval attempts (MESE, multifocal TESE, and microTESE) in 2010–2018 at the department of Andrology and Urology of FSBI National Medical Research Center for Obstetrics, Gynecology and Perinatology named after Academician V.I. Kulakov (Moscow, Russia) (n = 754). A subpopulation of 216 patients was formed. Patients had normal clinical-laboratory characteristics that included testicle volume, levels of testosterone, and gonadotropic hormones that were used for the preliminary diagnosis of OA.

The endocrine profile was evaluated with a Cobas e411 analyzer (Roche Diagnostics International Ltd., Rotkreuz, Switzerland) for immunochemical analysis. The level of inhibin B was measured with the Inhibin B Gen II test (Beckman Coulter United Kingdom Ltd., High Wycombe, UK). For a histological study, testicular tissue was fixed in Bouin solution and stained with hematoxylin and eosin. The degree of spermatogenesis disorders was assessed by the scale proposed by Bergmann and Kliesch (1998) [16] (Table 1). Furthermore, the data of the patients were compared with the results of the histological study. The authors compared preliminary and final diagnoses and established that some patients were reclassified from OA to NOA because of the signs of spermatogenesis disorder. All cases of diagnosis reclassification were further analyzed.

Surgical sperm retrieval techniques

MESA. Microsurgical Epididymal Sperm Aspiration was proposed by Temple-Smith et al. (1985) [17]. The surgery starts with a testis delivery via a 2.0 – 3.0 cm longitudinal incision. The epididymis membrane is dissected. Expanded tubules are found under a microscope and dissected. The appeared fluid is aspirated. MESA is repeated in various areas of the epididymis until the amount of sperm is obtained sufficient for fertilization and cryoconservation.

Multifocal TESE. Standard TESE surgery is performed without a microscope; still, a binocular loupe can be used. This procedure can be performed without complete delivery of the testes through the surgical wound. Instead, the manipulations are made via a small 2.0 cm window incision. Skin, fascia dartos, and tunica vaginalis testis are moved with a retractor, and tunica albuginea is dissected along 1.0 cm. Mechanical pressure on the testis allows a surgeon to obtain a fragment of the parenchyma via this incision. One of the most famous modifications of TESE is multifocal TESE when testicular tissue is probed from different sections of the testes [18-20].

MicroTESE. MicroTESE is performed via the same access as MESA. After testicular delivery, a median incision of tunica albuginea is made in the avascular area at x6 – x8 magnification. Dissection of the testicle parenchyma and search for areas with normal seminiferous tubules are performed at x16 – x25 magnification. Testicular tissue biopsy specimens are taken from these areas. If normal tubules are not found, during biopsy, any irregular tubules are excised. The wound is sutured layer by layer [21]. Out of all available methods of sperm retrieval, microdissection biopsy is the most efficient and safe one.

Statistical analysis. Statistical data processing was performed using the software package IBMÒ SPSS Statistics 23 (SPSS: IBM Company, IBM SPSS Corp., Armonk, NY, USA). The authors used multivariate regression analysis and decision tree analysis.

A total of 216 patients were preliminarily diagnosed with OA. However, in 131 (60.6%) patients, there were histological signs that indicated a spermatogenesis disorder (hypospermatogenesis, delayed sperm maturation, or Sertoli-cell-only syndrome (SCOS)). Twenty-two patients had true NOA associated with homogenous spermatogenesis failure at the stage of primary spermatocyte production. Sixty-three patients had histological signs of spermatogenesis disorders (hypospermatogenesis, late sperm maturation arrest, local SCOS pattern) associated with long-term obstruction of seminiferous ducts). Thirty-one patients had epididymoorchitis in their medical history. Iatrogenic disorders associated with long-term intake of antibacterial and hormonal drugs were suspected in 13 cases (Figure 1).

Multivariate regression analysis showed that only regular exposure to high temperatures (for example, hot baths, long-term stay near sources of high temperature) was an independent predictor of NOA diagnosis in patients with normal clinical and endocrine parameters (OR = 1.989; 95% CI = 1.101–3.595) (Table 2). It is suggested that hyperthermia did not act as an independent etiological factor of azoospermia but led to histological changes in patients with vulnerable spermatogenesis.

The largest group of reclassified cases included 63 patients with a clear etiology of long-term obstruction (>5 years) and a lack of additional risk factors. There were cases caused by various congenital pathologies (n = 23), bilateral epididymitis associated with sexually transmitted infection (STI) and not associated with STI (n = 18), surgery for bilateral inguinal hernia repair (n = 11), spermatocele removal (n = 4), vasectomy (n = 3), midline prostatic cysts (n = 3), and surgery for anorectal malformation (n = 1). Congenital anomalies were presented as Congenital Bilateral Absence of the Vas Deferens (CBAVD) (n = 10), partial unilateral anomalies of the mesonephric duct with seminal vesicle agenesis (n = 8), complete unilateral anomalies of the mesonephric duct with kidney agenesis (n = 2), Zinner’s syndrome (n = 2), and Young’s syndrome with transposition of internal organs (n = 1). The mean duration of obstruction in these patients was 16 years (interquartile range [IQR]: 10–24.5).

There are several clinical cases for better illustration of the observations.

Clinical case No. 1. Patient Ch., 32 years old. Spermogram: azoospermia with normal sperm volume. Family medical history: the patient’s sister died of pneumonia at the age of three years old. Physical examination: seminal ducts are not palpated. CFTR mutation status: negative. Hormonal status: testosterone — 10.2 nmpl/l, luteinizing hormone (LH) — 3.9 IU/l, FSH — 2.3 IU/l, prolactin — 112.0 mIU/l, estradiol — 106.0 pmol/l, Anti-Mullerian Hormone (AMH) — 3.9 ng/ml, inhibin B — 116.0 pg/ml. The patient was diagnosed with OA associated with the bilateral congenital absence of seminal ducts. During a surgical procedure of sperm retrieval, an epididymis with expanded tubules was revealed. Epididymal fluid had a high leukocyte count and some immotile sperms. Multifocal TESE revealed a significant amount of sperm for cryoconservation. A fragment of testicular tissue was sent for a histological study, which showed that in 45.0% of cases germ cell degeneration was degeneration of germ cells; 5.0% of tubules had histological signs of SCOS, and 50.0% of tubules did not even have Sertoli cells (“shadow tubules”). The elongated spermatids were not revealed (BKS = 0). It is suggested that the distribution of germ cells in testicles was uneven, with areas of preserved spermatogenesis (where sperms were revealed) and areas of impaired spermatogenesis (where the biopsy specimen was taken for a histological study).

Clinical case No. 2. Patient S., 67 years old. Spermogram: azoospermia with normal sperm volume. The patient has five children from a previous marriage. He underwent vasectomy 27 years ago. Hormone replacement therapy with testosterone was not performed. Hormonal status: testosterone — 18.2 nmol/l, LH — 4.1 IU/l, FSH — 6.3 IU/l, prolactin — 99.0 mIU/l, estradiol — 97.0 pmol/l, AMH — 2.2 ng/ml, inhibin B — 101 pg/ml. The couple applied for surgical retrieval of sperms with further ART. There was an attempt of MESA surgery, but it was impossible to identify the epididymal tubules due to total epididymal sclerosis. Multifocal TESE did not yield results. There was an attempt of its conversion into micro TESE, and a native cytological study revealed sperms in a testicle. Due to the inferior quality of the sperm, it was decided to retrieve the sperm also from the contralateral testicle, where the clinical picture was similar. Three straws of sperms were cryoconservated after the processing of samples obtained from two testicles during microTESE. A histological study of the right testicle showed that 20% of the tubules had SCOS, 60% of the tubules had spermatogonia and primary spermatocytes, and only 20% of the tubules had elongated spermatids (BKS =2). In the left testicles, 30% of tubules were “shadow tubules”, 60% of tubules had SCOS, and 10% of tubules contained spermatogonia (BKS = 0).

Acute bilateral epididymoorchitis in medical history was the only cause of obstruction in 30 patients. One patient had unilateral epididymoorchitis in medical history. Unlike epididymitis, epididymoorchitis was associated with fever, scrotal edema, pain in testicles, and palpation tenderness. Thirty-one patients underwent a successful sperm extraction, but in seven patients it was impossible to retrieve sperm from the epididymis due to total fibrosis of the epididymis; five patients required microTESE due to the insufficient number of sperm in testicles recovered after multifocal TESE. Post-inflammatory changes were clear during a histological study of testicle tissue, which indicated that spermatogenic insufficiency was secondary regarding testicle tissue inflammation in patients with acute epididymoorchitis.

Clinical case No. 3. Patient M., 37 years old. Spermogram: azoospermia with a normal volume of sperm. He had acute bilateral epididymoorchitis with scrotum edema, high temperature, and pain at the age of 13 years old in medical history. Physical examination: The right epididymis has uneven texture at palpation with a stiff head, atrophic body, and dense tail, painful at palpation. Hormonal status: testosterone — 17.3 nmpl/l, LH — 2.2 IU/l, FSH — 4.9 IU/l, prolactin — 230.0 mIU/l, estradiol — 89.0 pmol/l, AMH — 4.8 ng/ml, inhibin B — 75.9 pg/ml. An attempt of MESA was made on the right testicle; a significant volume of yellowish fluid was aspirated from the epididymis head. A native cytological study of epididymal fluid revealed a significant leucocyte count and no sperm. After unsuccessful multifocal TESE, microTESE was performed. The seminiferous tubules were uneven in diameter, color, and consistency. Two straws with sperm were thawed for further ICSI. A histological study showed that 53.0% of seminiferous tubules contained elongated spermatids, 3.0% of tubules had SCOS, and 44.0% of tubules were hyalinized and sclerosed (BKS = 5).

Thirteen patients underwent long-term treatment for male infertility (>12 months) before applying to the specified center for second expertise. Previously, such patients received empirical therapy that could negatively affect spermatogenesis, including any empirical hormonal stimulation, mainly clomiphene citrate and human chorionic gonadotropic hormone, which was prescribed without proper endocrine tests or follow-up. For example, a patient received 50 mg/day of clomiphene and 20 mg/day of tamoxifen for more than 6 months to stimulate spermatogenesis. Four of them received three four-week courses of levofloxacin within 12 months along with hormonal therapy before applying to the center. It is difficult to describe and classify empirical therapy, which was given to these 13 patients. It is known that these drugs were prescribed in unusual combinations and doses. However, all these cases had one common detail. The examination showed persistent obstruction of the seminiferous ducts. Two patients had prostatic cysts, two patients had inguinal hernia repair surgery, and nine patients had congenital abnormalities that were not detected during primary examination or ignored as clinically insignificant. It is not necessary to say that the appearance of sperm in the ejaculate would be unlikely after the prescribed pharmacotherapy in these cases. Sperms were retrieved from the epididymis in six cases. Multifocal TESE was successful in eight cases and conversion to microTESE was needed in five cases. In two patients, sperm retrieval was unsuccessful. It is suggested that unsuccessful microTESE was caused by the fact that, in both cases, there had been <20 months since the end of aggressive empirical therapy. The authors had to attempt sperm extraction without proper preparation because their female partners had to undergo an oocyte puncture.

Clinical case No. 4. Patient Ch., 31 years old. Spermogram: azoospermia with low or normal sperm volume (0.5–1.6 ml). Physical examination did not reveal any peculiarities. Medical history: the patient was observed by a urologist for two years and received antibacterial therapy for chronic prostatitis. Additionally, he received therapy with alpha-adrenoblockers, 25 mg clomiphene citrate, cabergoline and periodic injections of human chorionic gonadotropic hormone. Hormonal status: testosterone — 14.1 nmol/l, LH — 7.6 IU/l, FSH — 7.2 IU/l, prolactin — 204 mIU/l, estradiol — 113 pmol/l, AMH — 2.0 ng/ml, inhibin B — 96.1 pg/ml (baseline endocrine parameters are unknown). The examination included repeated transrectal ultrasound investigation, which revealed agenesis of the right seminal vesicle and the pelvic part of the seminal ducts. Ultrasound did not detect the right kidney and there was a suspected anomaly of a mesonephric duct. These data were verified by the results of magnetic resonance imaging. MESA and multifocal TESE were unsuccessful. One portion with testicular sperms was cryoconservated after microTESE. A histological study showed that 57.0% of seminiferous tubules contained elongated sperms, 19.0% of tubules had SCOS, and 24.0% — “shadow tubules” (BKS = 6).

All previously described groups of patients had seminal duct obstruction as the primary cause of infertility and aggravating factors that caused spermatogenic insufficiency. However, there were 22 patients with true NOA caused by a maturation cease at the stage of primary spermatocytes, which was shown by a histological study (BKS = 0). All patients required microTESE, which was successful only in two cases.

Clinical case No. 5. Patient B., 51 years old. Spermogram: azoospermia with normal sperm volume. Hormonal status: testosterone — 13.9 nmol/l, LH — 3.7 IU/l, FSH — 2.8 IU/l, prolactin — 253 mIU/l, estradiol — 76.4 pmol/l, AMH — 4.23 ng/ml, inhibin B — 200 pg/ml. The testicle volume was calculated by a formula: 21 cm3 (right testicle) and 22 cm3 (left testicle). Karyotype – 46, XY. No AZF mutations (Azoospermia factor) were revealed. The medical history was without peculiarities. There were no potential reasons for seminal ducts obstruction. During surgical retrieval of sperms, a hyperplastic epididymis was revealed without visible tubules and the MESA procedure was stopped. MicroTESE showed that all seminiferous tubules looked uniform, white, large, and firm. Sperms were not detected. A histological study showed a uniform delay of sperm maturation at the level of primary spermatocytes.

Decision tree analysis showed that patients with inhibin B < 93 pg/ml had the highest risk of incorrect primary diagnosis (82.6%) (Figure 2).

The present study revealed several clinical scenarios when differential diagnosis between OA and NOA becomes a complicated task. In most cases, obstruction was the main cause of infertility, but some factors caused secondary spermatogenic insufficiency, which was verified by a histological study. It is still disputable whether it is enough to reclassify long-term or complicated OA into NOA. It is suggested that NOA is not an adequate term for this condition because it can be interpreted as the absence of obstruction. There are no suitable terms to describe this condition. However, the authors believe that such reclassification is feasible because it affects the prognosis and therapy of the disease.

The number of retrieved sperms was lower than expected in patients with OA. For a significant number of patients, the conversion to microTESE was needed to find suitable sperms. Since not all reproductive health clinics are adequately equipped, further attempts at sperm retrieval should be planned, considering this fact.

Long-term obstruction of the seminal ducts can cause spermatogenesis disorders due to elevated pressure in the seminiferous tubules or immunological mechanisms. Testicular ischemia caused by a vascular lesion cannot be excluded as a separate factor in patients who underwent inguinal hernia repair or vasectomy. It is known that long-term obstruction negatively affects the restoration of semen ducts after vasectomy [22]. The study by Raleigh et al. (2004) demonstrated testicular fibrosis and impaired later stages of spermatogenesis associated with longer obstruction after vasectomy that could be caused by induction of apoptosis [14]. Spermatogonia, primary spermatocytes, and Sertoli cells might not have been affected, unlike in the study published earlier by Shiraishi et al. (2002) [23]. Considering this, patients should be advised on cryoconservation of sperms before vasectomy. The authors could not verify these histological data because of a small sampling of patients who underwent vasectomy because this method of contraception is not widespread in Russia. It is suggested that the correlation between long-term obstruction and secondary spermatogenic insufficiency was also true for other conditions, including congenital anomalies. Cito et al. (2019) published a clinical case of long-term obstruction with histological signs of SCOS in a patient with Zinner’s syndrome [13]. It was an example of consistency or an idiopathic case of NOA coexisting with Zinner’s syndrome.

The treatment of azoospermia in centers that do not specialize in reproductive andrology can worsen reproductive outcomes even if patients are further directed to a specialized department. A fertility-challenged couple can lose time trying to improve the spermogram parameters by conservative means of therapy. The authors observed 13 patients with congenital anomalies that cause seminal duct obstruction, which were not properly diagnosed preliminarily. The patients received aggressive therapy for over a year, expecting the appearance of sperms in ejaculate because of hormonal stimulation of resolution of suspected chronic prostatitis. However, it is known that such antibiotics as fluoroquinolones can negatively affect testicular tissue [24][25]. Drugs used for hormonal stimulation of spermatogenesis can also impose this risk. A paradoxical response to clomiphene therapy is well-documented in men without azoospermia, but it is hardly observed in men with azoospermia [26]. Human chorionic gonadotropic hormone is another popular medication prescribed for infertility, which can also cause spermatogenesis disorders because of a negative effect on FSH [27][28]. Thus, physicians must be aware of the underlying potential problems of endocrine-mediated therapy. The cases of these patients who initially received inadequate therapy highlight rarely discussed issues in this sphere.

The observations made by Dada et al. (2003) confirm that high intratesticular temperature causes partial or complete arrest of spermatogenesis and can lead to an increase in the production of morphologically anomalous sperms with impaired motility. This association between elevated temperature and sperm functioning plays a significant role in clinical medicine for the understanding of pathological conditions and therapeutic measures [29]. In the present study, the authors obtained data that confirmed the negative effect of regular high temperatures (for example hot baths, furnace) on spermatogenesis. It was an independent negative predictor of NOA in patients with normal clinical and endocrine parameters. The authors suggested that hyperthermia did not act as an independent etiological factor of azoospermia but led to histological alterations in patients with vulnerable spermatogenesis.

Differential diagnostics between OA and NOA are not always a simple task. Any case of azoospermia in patients with a normal volume of ejaculate can be classified by default as NOA if there are no amnestic or clinical data that would demonstrate obstruction (for example, vasectomy in medical history or absence of seminal ducts). Diagnostic challenges can be caused by secondary spermatogenesis disorders associated with long-term obstruction, epididymoorchitis in medical history, and total genetically caused arrest of spermatogenesis. It is suggested that regular exposure to high temperatures can provoke azoospermia in men with initially vulnerable spermatogenesis, which requires further studies. Finally, the reference values for inhibin B proposed by most laboratories are not suitable for evaluating reproductive function.