Differentiation between nonobstructive azoospermia and obstructive azoospermia: then and now

Abstract

Male infertility has seen an increase in prevalence with cases of azoospermia estimated to affect 10%–15% of infertile men. Confirmation of azoospermia subsequently necessitates an early causal differentiation between obstructive azoospermia (OA) and nonobstructive azoospermia (NOA). Although less common when compared to NOA, OA can represent upward 20%–40% of cases of azoospermia. While there are a multitude of etiologies responsible for causing NOA and OA, correctly distinguishing between the two types of azoospermia has profound implications in managing the infertile male. This review represents an amalgamation of the current guidelines and literature which will supply the reproductive physician with a diagnostic armamentarium to properly distinguish between NOA and OA, therefore providing the best possible care to the infertile couple.

INTRODUCTION

Male infertility continues to become increasingly prevalent in the population and cases of azoospermia are present in 10%–15% of infertile men.1,2 It has been estimated that based on the results of a recent meta-analysis, the prevalence of male infertility has increased 76.9% compared to the rate in 1990.3 Azoospermia, as defined by the World Health Organization (WHO), is the absence of spermatozoa in the sediment of a properly centrifuged ejaculate sample.4 When a patient presents with azoospermia, often the first determination a practicing reproductive urologist must make is whether he has obstructive azoospermia (OA) or nonobstructive azoospermia (NOA). OA is less common than NOA, but is still estimated to affect more than 20%–40% of men with azoospermia.5,6 There are a myriad of causes that may result in NOA or OA, and the consequent clinical management of both types of azoospermia can be heavily dependent on its cause. Typical findings of NOA and OA are summarized further in Table 1.

Table 1.

Distinguishing features of obstructive and nonobstructive azoospermia

Diagnostic findings and features	OA	NOA
Patient history	Painful ejaculation	Asymptomatic
	Hematospermia	Often painless
	Pancreatic insufficiency	Testicular infections
	Chronic sinopulmonary infections	Medications
	Seminal duct infections	Trauma and radiation
		Chemotherapy
		Opiates and alcohol
		TRT
Physical examination findings	Normal testicle volume	Testis volume <16 ml
	Absent or atretic Wolffian duct structures	Testis length <4.6 cm
	Dilated SV and epididymis
	Clinical epididymal nodule
	Midline prostatic cysts
Semen analysis	Low ejaculate volume azoospermia* ↓ Fructose pH <7	Normal ejaculate volume and azoospermia Normal pH >7
Laboratory testing	↔ FSH, LH, and inhibin B	Testicular
	↓/↔ Testosterone	↓/↔ Testosterone
	↔ Prolactin	↑ FSH >15 IU l−1#
		↔ LH
		Pretesticular
		↓ Testosterone
		↓ FSH
		↓ LH
		AIS
		↑ Testosterone
		↑ FSH
		↑ LH
Genetic testing	CFTR mutations: ∆508, 5T, 7T, and 9T	Karyotype (XXY, XO, and KAL) AZFa, AZFb, and AZFc mutations Translocation mutations (Robertsonian, and reciprocal) Inversion mutations
Imaging findings	Dilated epididymis Spermatocele Epididymitis Prostatic cysts EJD dilation >2 mm SV dilation >1.5 cm	Clinical varicocele >3 mm* Anterior pituitary pathology

^*Not a requisite finding for OA or NOA. ^#Normal range of laboratory values in the literature. SV: seminal vesicles; FSH: follicle-stimulating hormone; LH: luteinizing hormone; CFTR: cystic fibrosis transmembrane receptor; KAL: Kallmann syndrome; AZF: azoospermia factor; EJD: ejaculatory duct; TRT: testosterone replacement therapy; AIS: androgen insensitivity syndrome; ↑: high or elevated; ↓: low or decreased; ↔: normal or no change; OA: obstructive azoospermia; NOA: nonobstructive azoospermia

This review was generated by careful analysis and inclusion of major systematic reviews, professional and societal guidelines, randomized control trials, and well-done manuscripts relevant to NOA and OA. Given the rapidly increasing prevalence of male infertility at a population level, a growing number of men will inevitably present with complaints of infertility which will inexorably include azoospermia. A practicing reproductive physician will therefore frequently encounter men with azoospermia, requiring proper clinical identification and differentiation of NOA and OA. It is hence the purpose of this review to provide a compilation of laboratory, imaging, genetic and histologic features, as well as coming advancements in the field of azoospermia. This will in turn provide a basis to better distinguish NOA from OA to offer the best management for the infertile couple seeking care.

PATIENT HISTORY AND AZOOSPERMIA DIFFERENTIATION

As is the case with most presenting medical complaints, the generation of a working differential for azoospermic patients begins with a proper medical history. Patient-reported symptoms can be notoriously vague and often nonspecific; however, certain symptomatic complaints in the setting of infertility may lead a perceptive physician to consider one type of azoospermia over another. NOA can be a physically painless phenomenon with the presenting male at times being asymptomatic. However, men with NOA also often have concurrent hypogonadism. In a recent study of 767 men with NOA, the overall period-prevalence of biochemical hypogonadism was estimated at 80.8% (95% confidence interval [CI]: 77.9%–83.4%).7 Furthermore, concomitant primary or secondary testicular failure in young men <40 years old can present with resultant signs and symptoms of hypogonadism even at traditionally “higher” levels <400 ng dl⁻¹ as evidenced by Scovell et al.8

Similarly, most men with OA have no specific complaints. However, men with congenital bilateral absence of vas deferens or with ejaculatory duct obstruction (EDO) may present with low volume ejaculate.9 Furthermore, those men with EDO may complain of ejaculatory pain and less commonly with the “classic triad” of painful ejaculation, infertility, and hematospermia.9,10,11

A carefully performed medical history of the infertile couple is vital to a proper azoospermia workup. This may reveal risk factors which can prove helpful in determining whether a patient has NOA or OA. Historical features can not only be used to differentiate NOA from OA but also indicate the level of obstruction. A history of genitourinary (GU) infections should be elicited in addition to a proper sexual history and screening for sexually transmitted infections. Infections can result in obstruction at essentially any level of the reproductive tract and is important to be aware of when attempting to differentiate NOA from OA.12

Aspects of a patient’s history such as testicular torsion, trauma, and cryptorchidism can be found both in OA and NOA. The rates of development of NOA are variable among these conditions and are largely dependent on the functional status of the contralateral testicle. Traditionally, unilaterally cryptorchid men have a fertility rate of 89.5%, which approaches the general population rate of 94%. However, boys with bilateral cryptorchid testes carry an NOA rate of approximately 40% and a six-fold increase in the risk of infertility.13 Similarly, an NOA incidence of 30%–40% has been reported in males with torsion and an abnormal contralateral testicle.13,14 A history of pelvic and scrotal trauma is an uncommon feature in those with OA; however, a much more common cause of OA stems from iatrogenic injuries. A thorough and complete surgical history may help differentiate the two variations of azoospermia. A large retrospective study of 324 men undergoing vasoepididymostomy for OA found that almost 20% of men in the cohort had iatrogenic injury as a causative factor. Additionally, 22% of men had infection as a primary etiology, with only 1.5% of men having a history of trauma.15

The type of surgical procedure that was performed in the case of iatrogenic injuries can inform a physician about the type of expected OA. During an OA workup, a history of scrotal trauma and/or surgery (such as hydrocelectomy, epididymal surgery/injury) may be encountered.16 In a cohort of men, it was discovered that approximately 7% of men being explored for OA had an iatrogenic injury to the vas deferens as an inciting event. The rates of vasal obstruction, however, are variable in the literature and dependent on the type of surgery (hernia repair and renal transplant being the most common), timing of the surgery, approach, and even the type of mesh utilized.11,17 Epididymal obstruction must remain on the differential in situations of prior iatrogenic vasal injury as well. Chronic epididymal obstruction can frequently cause a buildup of epididymal luminal pressures causing rupture injury to the tubules in up to 60% of men.6,18

A history of gonadal toxins, radiation, and various pituitary lesions should prompt one to consider NOA. However, uniquely, a history of neurologic symptoms or surgery/trauma involving the pelvic nervous system should strongly implicate a functional neurogenic obstruction of the distal seminal ducts. Local neurogenic and urodynamic dysfunction has been shown on vesiculography to be an infrequent cause (in 6 of 94 men evaluated for OA) of OA in prior studies.19 Other historical features such as multiple sclerosis, spinal cord injuries, pelvic neurologic trauma or surgery, and neurological injury during retroperitoneal lymph node dissections can all cause functional obstructions due to a damaged mechanism of semen transport through the genitourinary tract.20 Although these are rare causes of OA, they remain important to keep on a complete differential when considering a patient with azoospermia.

Genetic conditions can present with a wide array of symptoms; however, there are certain features that are common to specific conditions that a vigilant physician may utilize to discriminate between the different types of azoospermia. Syndromic NOA is most frequently associated with Kallmann syndrome (KAL) and Klinefelter’s syndrome (KS). Patients presenting with the constellation of infertility, atrophic testicles with underdeveloped secondary sexual traits, and especially and specifically anosmia or hyposmia should be screened for KAL.21 While KS can involve a range of signs and symptoms, patients will often present with some or all of a classic pattern of atrophic testis, gynecomastia, a eunuchoid appearing phenotype with infertility, and learning difficulties.22 Men presenting with a history of chronic sinopulmonary infections and pancreatic insufficiency deserve special mention. These signs and symptoms in the setting of OA may signal a congenital/genetic obstruction due to cystic fibrosis (CF) or Young’s syndrome.23,24

PHYSICAL EXAMINATION

While a careful medical history is vital to a proper azoospermia workup, male infertility is rather unique in urology, given that physical examination can at times be essential in the diagnosis of OA versus NOA. Although an examination in a patient with NOA may often be within normal limits, there are certain features present on a physical examination that more strongly suggest NOA as a cause for fertility issues. A thorough scrotal examination may reveal a varicocele, which has been demonstrated to be associated with infertility and occasionally NOA. Several large studies including recent meta-analyses have shown that microsurgical repair leads to an improvement of abnormal semen parameters, and at times promotes the reappearance of sperm in the ejaculate of patients with NOA prior to surgical sperm retrieval.25,26 Appropriate sexual development is important to note, as undervirilization, diminished body hair, and gynecomastia with a eunuchoid phenotype can indicate androgen deficiency. This constellation of symptoms can exist either in the presence or absence of syndromic etiologies of NOA as well.27 Men with OA conversely are in general phenotypically well virilized without the aforementioned signs or symptoms.

Examination of the testicles and the accessory structures merits special discussion in regard to the topic of azoospermia. Testicular volume must at minimum be estimated or more accurately quantified using a Prader orchidometer. It has been estimated that around 400–600 seminiferous tubules are found within the testis, representing approximately 80%–90% of the total testicular volume; therefore, decreased testicular volume and soft consistency logically correlate to decreased spermatogenic potential. Total testicular volumes >15 ml tend to be more indicative of OA. Conversely, as testicular volume declines, there is an almost linear worsening of sperm parameters (including increasing DNA fragmentation, mitochondrial changes, and chromatin abnormalities), with a testicular volume <2 ml being the point at which testicular function precipitously declines.28 Using an orchidometer, it has been well established that normal testicular length should be >4.6 cm with a width of 2.6 cm. Testicular long axis measurements of <4.6 cm predict spermatogenic failure and NOA as a cause of azoospermia almost 90% of the time. However, recent evidence also indicates that higher follicle-stimulating hormone (FSH) and smaller testicular volumes are strongly associated with more severe histologic patterns of NOA.29,30 Inversely, over 96% of men with OA were found to have FSH <7.6 mIU ml⁻¹ and testicular length >4.6 cm, indicating the utility of the physical examination findings in the differentiation of NOA and OA.29

Systematic examination of the testicular accessory glands is also crucial in evaluating NOA and OA. Starting with the vas deferens, bilateral or unilateral absence of the vasa may be a harbinger of CF mutations and merits genetic testing. Although congenital bilateral absence of the vas deferens (CBAVD) is archetypal of CF, there can be additional Wolffian duct abnormalities with either epididymal obstruction or upstream vasal obstruction being commonplace.31 This type of OA often presents with an intact caput of the epididymis, normal testicular size and consistency, and absent or hypoplastic seminal vesicles (SV).32 Urogenital abnormalities are common in men with CBAVD and unilateral absence of the vas. A unilateral absence of the vas deferens commonly presents with seminal duct abnormalities and a rate of renal agenesis of 26%. Conversely, men with CBAVD have renal agenesis rates of 11%.33

Overall, the most common cause of acquired OA remains elective sterilization in the form of vasectomy. This will present as a small defect of a vasal segment usually bilaterally. Often, clips may be palpated and delineate the area of obstruction. Concurrently, there may also be epididymal obstruction from back pressure in vasectomized men. This often takes the form of a palpable nodule due to intravasal hyperpressure or a clinical epididymal nodule at the end of the proximal vasal stump indicating potential OA.34 While the diagnostic yield is admittedly low, a digital rectal examination should be considered in certain settings when EDO is being considered in an infertile male. On rare occasions, EDO may be diagnosed on a rectal examination by palpation of cystic dilations of the ejaculatory ducts or of a cyst along the midline of the prostate. Furthermore, the seminal vesicles should never be easily palpable, especially after a short period of sexual abstinence. Their prominence on examination should always alert to OA as a potential cause of infertility.35

SEMEN ANALYSIS

Potentially, the most critical aspect of an azoospermia workup involves semen analysis. Azoospermia is defined as the complete absence of spermatozoa in the ejaculate. The diagnosis is confirmed after centrifugation and microscopic examination of the pellet.4 Unilateral or partial obstructions of the vas or epididymis may produce oligospermia or cryptozoospermia; however, with a normal contralateral vas, the rates of azoospermia are essentially zero. In fact, obstructions proximal (or upstream) to the level of the ejaculatory ducts tend to have normal seminal volume (>1.5 ml) due to the fact that over 90% of the semen volume is derived from the contributions of the prostate and seminal vesicles.36 Yet, despite this physiologic sequitur, ejaculate volume is seldomly a reliable diagnostic feature of OA. In NOA, the primary pathologic process for diminished spermatogenesis is often primary testicular failure, in some cases caused by hypothalamic–pituitary–testicular axis dysregulation. Due to this fact, semen analyses in men with NOA can have a normal ejaculate volume, but lack of spermatozoa on formal analysis.37

Therefore, it begs the question: are there other features of the semen analysis that can differentiate NOA from OA? Normal semen is quite cellular and alkaline at baseline. In situations of OA, the sample trends toward a more translucent consistency due to a loss of sperm count, overall cellularity, and increased liquefaction. These findings are a result of predominantly prostatic secretions in the case of unilateral seminal tract obstruction and a watery ejaculate in cases of EDO or CBAVD.36,38,39 Thus, physiologic obstruction of the ejaculatory ducts or seminal vesicle atresia as seen commonly in CF patients with CBAVD, will lack the alkaline proximal seminal tract contributions. This results in the deposition of primarily acidic prostatic secretions and generates a characteristically low seminal pH <7 or <6.8 in CBAVD. Finally, although its diagnostic utility has fallen out of favor in recent years, semen fructose can be measured in cases of OA. Fructose is primarily produced in the seminal vesicles; therefore, with EDO, a positive finding on resorcinol reagent testing can point toward obstruction as the cause of azoospermia.40

Another feature that has been explored for its role in differentiating OA from NOA is the presence of immature germ cells (IGCs). These cells often coalesce on a semen analysis as “round cells” which include IGCs and various white blood cells. It has been postulated that elevated IGCs are present in situations or conditions where sperm integrity and quality has been affected. Instances where the testicular microenvironment has been disrupted (as in varicoceles, spermatogenic toxins, or maturation arrest) can therefore result in shedding of immature germ cells.41 This logic has been parlayed into testing for OA. In theory, with OA, there is intact spermatogenesis and a total obstruction of the seminal pathways; therefore, there will be no shedding of IGCs, and staining of the pellet in OA should not identify IGCs.42 However, these immature cells can be difficult to differentiate from degenerating inflammatory cells and are not always helpful in diagnosing NOA. Due to these limitations, the current WHO considers testing for semen leukocytes and antibodies part of an optional extended evaluation and no longer part of the standard semen analysis testing.4

There are several methods that may be utilized in identifying seminal germ cells. Amer et al.43 obtained seminal fluid from 100 men with NOA and performed a May-Grünwald stain in order to identify round spermatids in the semen. Of these men, sperm was detected in 87.3% of cases, while sperm was found in only 22% of testicular sperm extraction (TESE)-negative cases, offering an interesting and resource efficient way of detecting potential spermatogenesis in men with NOA.43 Flow cytometry has also emerged as an alternative technique in finding and quantifying seminal round spermatids in order to identify testicles that may harbor intact spermatogenesis. In a study of 37 biopsies obtained from men with NOA, flow cytometry was found to be more sensitive (100% vs 59%; P = 0.0001), but less specific (67% vs 83.5%; P = 0.0034) when compared to traditional cytology. Therefore, a combination of the two modalities may provide an alternative to the histologic sampling which was specific (100%) but less sensitive (50%) in predicting patients with NOA harboring intact spermatogenesis.44 Similar principles have been utilized looking at the ploidy of IGCs. Yeung et al.45 conducted a study in which the ejaculates of 50 infertile males were analyzed based on a fluorescent mitochondrial marker to select out high-ploidy contaminants. They found no flow cytometry peak of highly condensed chromatin (indicating the presence of IGC products) in the spermatozoa in men with OA when compared to other NOA and idiopathic ejaculates.45 These technologies offer improved IGC detection and can offer valuable information in differentiating NOA from OA; however, they are difficult to apply with patients in clinical practice which has somewhat limited their widespread use.

Other studies have looked at the role of specialized centrifugation and staining techniques to additionally help determine if rare sperm are present in an azoospermic sample. Sharma et al.46 used cytocentrifugation and staining in combination with nuclear fast Picroindigocarmine staining to help identify sperm in azoospermic samples. Semen analysis was performed on 251 men with diagnosed azoospermia, of which 60/251 men (23.9%) had sperm detected on a cytocentrifugation and staining.46 Although the data are retrospective in nature, the results are promising and the technique is relatively simple to perform. However, further validation studies should be conducted prior to being implemented on a larger scale.

HORMONAL AND GENETIC TESTING

A properly collected semen analysis is the foundation for a proper infertility workup. However, despite all the previously described diagnostic criteria, a formal diagnosis always requires hormonal confirmation. As per societal guideline recommendations, all men with azoospermia should complete a hormonal evaluation. In fact, a recent large cohort study found an approximately 20% incidence of hypogonadism in men with infertility. The same group additionally found that primary and compensated hypogonadism carried a multifold worse prognosis of infertility, smaller testicular volume, and azoospermia when compared to eugonadal men.47 This truly highlights the importance of properly interpreting hormonal workups, which in general includes FSH, luteinizing hormone (LH), prolactin (PRL), and total testosterone. Although less commonly used, inhibin-B testing can also be considered. An emerging biomarker in the setting of azoospermia is 17-hydroxyprogesterone (17-OHP). Several studies have shown a correlation between 17-OHP and intratesticular testosterone levels.48,49 A recent study was able to demonstrate an improvement in intratesticular testosterone (using 17-OHP as a surrogate) with human chorionic gonadotropin (hCG) and clomiphene citrate treatments over patients on testosterone replacement and healthy fertile controls (P < 0.05).48,49 This may play an important role in potentially distinguishing azoospermic patients (especially men with NOA) who may benefit from hormonal therapy before microdissection testicular sperm extraction (micro-TESE) and sperm retrieval, such as those with azoospermia and low intratesticular testosterone.48

Regarding OA, certain laboratory thresholds are sensitive for predicting OA. Since spermatogenic activity and Sertoli/Leydig cell function remains intact, one expects the FSH to be within normal limits and less than 7.6 IU l⁻¹. Similar to FSH, inhibin-B is also a Sertoli cell hormone product secreted apically into the seminiferous tubules.50 Inhibin-B has therefore been studied as a proxy for Sertoli cell health and spermatogenic function and is postulated to be normal in men with OA compared to NOA. Inhibin-B and FSH both reliably discriminate OA from NOA; however, the diagnostic yield for inhibin was not found to be better than FSH alone, and neither was predictive of sperm retrieval rates.51 Although FSH adds significant diagnostic value, it is important to note that a normal FSH does not always correlate to OA in every case. There are instances where concomitant NOA may be present in the setting of a normal FSH, such as spermatogenic arrest and Sertoli cell-only syndrome (SCOS).

An additional serum and seminal hormone that has been studied as a marker of spermatogenesis in NOA is anti-Müllerian hormone (AMH). AMH is part of the transcription growth factor beta (TGF-β) family and is preferentially secreted from Sertoli cells within the testicle. Historically, studies have found that AMH was lower in subfertile men when compared to normal controls. A recent study in 155 azoospermic men (23 with OA and 132 with NOA) found that serum AMH levels were markedly lower across all etiologies of NOA compared to OA with the lowest values in cases of genetic NOA (median: 7.3 pmol l⁻¹; P < 0.0001) and nonmosaic KS (median: 2.3 pmol l⁻¹; P < 0.0001).52 Recent research has tried to expand this application by looking at the AMH:testosterone ratio as a positive predictor of sperm extraction results on TESE with some promising results.53 However, there is conflicting data regarding AMH’s utility as a spermatogenic marker including a meta-analysis, which concluded that there was inadequate evidence in support of seminal or serum AMH; thus, more evidence is required before mainstream use is justified.54

In men with either CBAVD, or any degree of Wolffian-related structure hypoplasia, strong consideration should be given to genetic testing for CF mutations. The most common mutations resulting in the CF phenotype involve alterations to the cystic fibrosis transmembrane conductance regulator (CFTR) gene encoding a ubiquitous chloride channel.55 There are truly a plethora of mutations that result in the CF phenotype including rare mutations and polymorphisms that require a high diagnostic index of suspicion. In fact, it is estimated that up to 78% of men with CBAVD or unilateral absence of the vas will harbor a minimum of one CFTR mutation. CF mutations are common, especially with respect to CBAVD, and include ∆508, 5T, 7T, and 9T mutations.56,57 These mutations are important to identify because despite them causing obstruction-related infertility, spermatogenic function largely remains intact. This allows couples the potential for pregnancy through the use of assisted reproductive technologies. A large study found that from a cohort of 1347 males clinically affected by CF, the overall fertility treatment rate was low at 1% with pregnancy rates in the overall cohort around 1.3%; however, the live term birth rate among male partners was high at 67%.58 The same study additionally found that while poor respiratory health was a predictor of reduced fertility in females, respiratory status does not appear to affect reproductive outcomes in male patients.58 While there are other syndromes thought to be related to OA, such as Young’s syndrome, no current standardized genetic testing is known to be diagnostic, and genetic testing is limited to CF at the moment.

NOA presents an entirely different diagnostic challenge. The same hormonal evaluation is mandatory, but the interpretation is wholly diametric. NOA includes pretesticular and testicular causes. As opposed to OA, prolactin may be important and its elevation can raise suspicion for NOA stemming from a range of pathologies including prolactin secreting lesions, dopaminergic medications, or even opiate abuse.59,60,61 In secondary hypogonadism, NOA results from hypothalamic or pituitary dysfunction. This in turn leads to a failure to produce FSH and LH at normal levels, which consequently diminishes intratesticular testosterone and testicular function.62 NOA ensuing from primary testicular failure is very characteristic on laboratory testing with findings including elevated levels of LH and FSH, of which the latter generally has laboratory values >7.6 IU l⁻¹.29

All men found to have NOA should undergo genetic testing and counseling. As technology has advanced, the amount of specific genetic abnormalities that can lead to NOA is truly astounding. In brief, it can be broken down into sex chromosomal anomalies, autosomal abnormalities, and microdeletions. Sex chromosome abnormalities are common and include conditions such as KS. Autosomal mutations range from chromosomal translocations to severe inherited disorders with significant phenotypic anomalies and NOA such as Prader–Willi.63 It is for this reason that a mandatory aspect of genetic testing in NOA involves a standard karyotype. Partial and complete microdeletions of the Y chromosome are also of importance in NOA. Azoospermia factors (AZF) are areas of the Y chromosome long arm (Yq) responsible for controlling spermatogenesis. Complete and partial deletions of AZFa, AZFb, AZFb+c, and AZFa+b+c (as well as the more recently discovered partial deletions of g1/g2, r1/r3, r1/r4, and r2/r4 [gr/gr] mutations) are vital to screen for in men with NOA. This testing has ramifications for genetic and family planning counseling, primarily because complete deletions of the aforementioned regions harbor no chance of surgical sperm retrieval on micro-TESE.64 Traditionally, a sperm concentration below 5 × 10⁶ ml⁻¹ was considered a trigger for genetic testing. However, recent research has found that the diagnostic yield of these genetic tests is significantly diminished in patients with sperm concentrations >1 × 10⁶ ml⁻¹.65 Therefore, some level of prudence may be justified in ordering these genetic tests in NOA, and they should not be utilized in men with confirmed isolated OA.

IMAGING FINDINGS

Often, the diagnosis of OA versus NOA is not clear, and additional imaging and testing may be required. In the past, the only true recourse in diagnosing an obstructive process along the distal portion of the seminal tract was vasography, though now with advances in transrectal ultrasound, its use is considered primarily historical. When performed properly, it allows for the accurate visualization of the seminal tract from the vas deferens all the way to the ejaculatory ducts confirming patency. However, it is not without risks; vasography can generate scar tissue and therefore obstruction to either the vas or epididymis (if backflow is present).66 These limitations, and the improvements of other modalities, have relegated vasography to mainly historical interest in regard to differentiating NOA from OA.

In addition to physical examination, perhaps, the imaging modality that offers the most value as a diagnostic test is ultrasonography. Scrotal ultrasonography is one of the modern mainstays of an infertility workup as an adjunct to physical examination when indicated. Ultrasound of the scrotum and its internal contents can complement proper history, laboratories, and examination with valuable information about possible etiologies of OA versus NOA. Standard scrotal ultrasound involves duplex imaging of the testicle, epididymis, and spermatic cord in addition to estimations of testicular volume.67 Ultrasound findings in patients with NOA can often be more subtle than those found in OA as evidenced by a study from Abdulwahed et al.68 They found that scrotal ultrasound was more sensitive in detecting NOA (75%), but more specific in detecting OA (87%). Furthermore, transrectal ultrasound (TRUS) was more sensitive than scrotal ultrasound for OA, and both imaging tests had a better sensitivity and a better specificity for obstructive etiologies (45% and 83%, respectively), making it more ideally suited for cases of OA.68

Scrotal ultrasound can be helpful in distinguishing OA from NOA in several ways. While examination can suggest the diagnosis of OA under many circumstances, there are situations where physical examination becomes unreliable. For instance, large hydroceles, lymphedema, or small and inguinal testicles can obfuscate underlying anatomy. Scrotal ultrasound can therefore aid in finding testicles and associated structures in addition to delivering a more accurate anatomic description. Scrotal ultrasound may also help distinguish several different classes of obstruction with a high level of sensitivity and specificity. In general, as previously described, OA has a more normal testicle volume and size. Dilated epididymes, spermatoceles, or epididymitis can point to an epididymal level of obstruction. Chronic epididymitis is an important condition to highlight as its diagnosis has significant implications to OA especially. Imaging findings in chronic epididymitis can at times be normal; however, Doppler ultrasound can provide information on vascularity, echogenic heterogeneity (e.g., hyperechoic scar) in an otherwise edematous and indurated epididymis.69,70 While these findings alone do not always translate directly into OA, they can be important sonographic precursors to developing pathology. Defects in the vas deferens, or bilaterally nonpalpable or uncertainly palpable vasa can indicate a vasal level of obstruction. In addition, an ectasic rete testis or dilated efferent ductules may be further indicators of downstream obstruction.68,71

Scrotal ultrasound can also provide valuable adjunct information in terms of diagnosing NOA as well. Testicular volume in men with NOA tends to be smaller than those with OA. As testicles develop, seminiferous tubules mature and the level of echogenicity on ultrasound increases. Scrotal ultrasound may show indications of testicular dysgenesis, including mild-to-moderate inhomogeneity suggestive of impaired testicular function, though admittedly, its specificity is low in this regard. Severe nonhomogeneous echotexture on ultrasonography can also warn of testicular malignancy developing or present in the area which has been associated with NOA in 1%–3% of severe male factor infertility patients.72,73 In a large study of 1787 men with infertility, 451 of which had azoospermia, there were 29 cases of cancer which was 16.7 times the expected population rate. Men with azoospermia had a 2.2-fold higher risk of finding cancer versus infertile men who were not azoospermic.74 Furthermore, some authors have looked at the size of seminiferous tubules on scrotal duplex ultrasound undergoing micro-TESE. Nariyoshi et al.75 found that men with larger-size testes/tubules (≥250 μm) had a sensitivity and a specificity of 77% and 81%, respectively (area under the curve [AUC]: 0.82) of predicting sperm retrieval rates on TESE. While this still requires further validation, it represents interesting findings on standard ultrasonography that may aid in the differentiation and the subsequent management of the most complicated cases in male infertility.

Scrotal ultrasound may also help in distinguishing clinically significant varicoceles that may benefit from varicocele repair. On ultrasound, varicoceles whose greatest diameter is ≥3 mm during Valsalva in the upright position, as well as reversal of venous flow (reflux) ≥2 s have been found to be clinically significant. These ultrasonographic features may help decision-making as to whether to offer corrective treatments for an infertile male, especially in the setting of NOA.76,77 However, the American Urological Association/American Society for Reproductive Medicine (AUA/ASRM) guidelines do not recommend scrotal ultrasound in the male infertility evaluation unless the physical examination is difficult or diagnosis is equivocal.78,79,80 Another important imaging modality used to discern OA versus NOA is transrectal ultrasonography. TRUS is particularly well suited for the diagnosis of OA, especially when due to EDO. If OA is on the differential, TRUS can provide vital information regarding cystic obstructions, calcifications, seminal vesicle dilation, or anatomic variations such as Müllerian duct anomalies that can cause obstruction along the seminal tract.81 EDO is particularly important to recognize in the diagnosis of OA, often with findings of seminal vesicle diameter ≥1.5 cm and/or ejaculatory duct diameter ≥2 mm.66,82 However, there are several drawbacks to TRUS imaging. Dilated seminal vesicles are a sensitive but not specific sign of EDO in OA; furthermore, men without OA and EDO can have dilated seminal vesicles depending on their state of abstinence. A recent study correlated SV diameter and area with abstinence finding a correlation between both variables with increased duration of abstinence (r = 0.372, P = 0.0001).83,84 Therefore, it seems that ultrasound is an imperfect diagnostic tool, which has led to the exploration of magnetic resonance imaging (MRI) in infertility cases due to its enhanced tissue resolution.

As MRI technology has improved since its conception, its utilization has markedly increased throughout broad swaths of the medical field including male fertility. In a similar manner to TRUS, MRI can show anatomic variation, prostatic cysts, and seminal dilations with increased tissue resolution identifying 23% more aberrations over TRUS.82 However, this modality still suffers from similar drawbacks as TRUS. It has an inability to correlate an obstructive process to observed anatomy or clinical scenarios. Additionally, MRI lacks the ability to provide functional real-time images. Given this reality, recent research has been devoted to functional MRI (fMRI) as a means to differentiate OA from NOA. Diffusion-weighted imaging (DWI) has shown promise in potentially distinguishing OA from NOA. Normal cellular structures (Sertoli cells, Leydig cells, and lymphatic channels), as in normal men or OA, tend to restrict the flow of water, whereas damaged or absent spermatogenic structures allow the free diffusion of water in the area correlating to NOA.85 A recent single-center trial found that an increase in the degree of proton water restriction (apparent diffusion coefficient or ADC) above 0.465 had a predicted sperm recovery rate, meaning that normal ADC values are associated with a higher predicted positive TESE in NOA.86 Emerging data on the use of MRI spectroscopy (MRS) have also provided insight in the identification of spermatogenesis and spermatogenic pockets in men with NOA. Choline peaks have been looked at as a marker for sperm membrane integrity in several studies due to its association with lipid metabolism.87,88 Increased choline peaks on MRS have been associated with areas of the testicle that still may harbor spermatogenic function. Conversely, areas with little to no choline activity imply decreased cellular function and turnover (hypospermatogenesis/SCOS).87 This may offer a potential “heat” map prior to sperm retrieval in men with NOA and may also add value in distinguishing NOA from OA.89 In fact, small pilot studies have shown an improved sperm retrieval rates in men with high choline detected on MRS prior to micro-TESE, with a cut-off value of 1.24 parts-per-million (ppm) predicting positive sperm retrieval (AUC: 0.665, P = 0.01; 95% CI: 0.722–1.00).90 Although these results are exciting, further study and confirmation are required.

The utility of MRI in infertility does not just extend to imaging of the testicles and lower genitourinary tract. Men presenting with infertility and hypogonadotropic hypogonadism (especially in the setting of elevated prolactin and suppressed gonadotropins and testosterone) should raise suspicion for a space-occupying pituitary lesion. According to society guidelines, a pituitary MRI is indicated in instances of hyperprolactinemia on at least two separate laboratory tests, testosterone <150 ng dl⁻¹ and low gonadotropins.91 MRI offers superior soft tissue definition important to cranial imaging, and its sensitivity and specificity are correlated to the size of the pituitary lesion, and the prolactin levels in the case of prolactinomas. However, MRI is still an imperfect study and attention has been paid to maximizing the diagnostic potential of pituitary MRI in the setting of infertility. Looking at the prolactin-to-testosterone ratio, values >0.1 with a prolactin cut-off of 25 ng ml⁻¹ were able to achieve 90% sensitivity and 48% specificity while reducing overall diagnostic costs.92,93

TESTICULAR BIOPSY AND ADJUNCTIVE TESTING

As recently as 2010, many societal guidelines indicated testicular biopsy as the best procedure to generate confirmatory histologic diagnosis in the setting of OA or NOA and was hailed as an important prognostic factor in sperm retrieval rates.94 Biopsies can be of use in situations where there is normal testicular volume on examination in conjunction with borderline FSH, a scenario which can be problematic in differentiating OA and NOA. International guidelines now recommend that testicular biopsy not be performed in the absence of therapeutic sperm retrieval to avoid subjecting patients to multiple invasive procedures.95 Biopsy in the setting of OA will mostly reveal normal histology. However, a testicular biopsy in men with NOA may reveal a number of different histopathologies including hypospermatogenesis, maturation arrest, SCOS, and even tubular hyalinosis. Intraoperative testicular biopsies, thus, allow for the assessment of sperm quality in addition to quantity, as well as germ cell neoplasia in situ, an underlying diagnosis not to be overlooked in patients with NOA.96 Despite providing a detailed cellular window into spermatogenesis, there has been little correlation between the above pathologic determinations and the ability to predict spermatogenesis and sperm retrieval in men with NOA.97 A more recent study showed that the overall sperm retrieval rate was associated closely with testicular histopathologic heterogeneity. A 33% sperm retrieval rate was found when a single histologic subtype was present versus 94% when four subtypes were present. This finding implies that multiple subtypes of histology (including presumably almost normal spermatogenesis) are more likely to yield sperm on extraction.98 Another study suggested that men with hypospermatogenesis histopathologic patterns had higher sperm retrieval rates, clinical pregnancy rates, and live birth rates than those with more severe histopathological architecture.99 This demonstrates that patients should be aware of the limitations of biopsy pathology and the need for future multidisciplinary studies in this area.

More recently, there has been a major push in infertility research to find serum or seminal biomarkers that can help with the diagnosis of different types of azoospermia, as well as the likelihood of finding sperm on micro-TESE. Testis-expressed gene 101 (TEX101) is a cell-membrane protein expressed singularly in testicular cells that is shed into the seminal plasma. Studies in this protein have yielded exciting results in the diagnostic dilemma that is azoospermia. Korbakis et al.100 compared different groups of patients (healthy fertile prevasectomy men, and men with unexplained infertility, oligospermia, and azoospermia) and found that patients with SCOS and OA had undetectable levels of TEX101. Moreover, they found that using a cut-off value of 0.9 ng ml⁻¹ for TEX101 generated a 100% sensitivity and a 100% specificity for distinguishing pre- and postvasectomy men.100 In the same study using enzyme-linked immunosorbent assay (ELISA) in combination with an epididymis specific protein extracellular matrix protein 1 (ECM1), they were able to distinguish NOA from OA with 81% sensitivity and 100% specificity. There are numerous other biomarkers that are being studied as they relate to the management of NOA and OA. Some that bear mentioning are transketolase like 1 (TKTL1), lactate dehydrogenase C (LDHC), and phosphoglycerate kinase 2 (PGK2). It has been demonstrated that low levels of the PGK2 marker is indicative of impaired spermatogenesis.101 Prostaglandin D2 synthase (L-PGDS) is also expressed in Sertoli cells and its levels are in general undetectable in the presence of OA such as vasectomized patients.102,103 While all these results are exciting, none are currently ready for mass use. However, they remain a promising step forward for infertility physicians in what remains to this day a diagnostic conundrum.

CONCLUDING REMARKS

Differentiating between OA and NOA has evolved over time, but this determination remains the most important step in evaluating the azoospermic male. Patient history, physical examination, and laboratory testing can help make this distinction in most cases. While imaging studies do not play a central role in infertility workup, they can be employed in diagnosing certain etiologies of OA and NOA. If a diagnosis remains unclear, a testicular biopsy remains a useful tool to help make the diagnosis when performed simultaneously with sperm retrieval. With advancements in the genetics of male infertility and next-generation sequencing, several biomarkers are being evaluated along with improvements in imaging modalities which will help in the assessment of azoospermia in the future.

AUTHOR CONTRIBUTIONS

LH provided the conception and design of the study and analysis and interpretation of the included studies, and drafted the manuscript. AR contributed to the interpretation of available studies and critical revision of the manuscript. GC provided conception of the design of the study, critical revision of the manuscript, and final approval for submission. All authors read and approved the final manuscript.

COMPETING INTERESTS

All authors declare no competing interests.