Testicular histopathology as a predictor of a positive sperm retrieval in men with non-obstructive azoospermia

Authors


Correspondence: David Ralph, The Institute of Urology, University College London Hospital, 235 Euston Road, London NW1 2BU, UK.

e-mail: dralph@andrology.co.uk

Abstract

What's known on the subject? and What does the study add?

  • The management of patients with non-obstructive azoospermia (NOA) involves testicular sperm extraction (TESE or microdissection TESE) combined with intracytoplasmic sperm injection (ICSI). Sperm retrieval is successful in up to 50% of men with NOA; however, there is no single clinical finding or investigation that can accurately predict a positive outcome. Several studies have concluded that testicular biopsy is the best predictor of a successful TESE.
  • The present study shows that the strongest predictor of the success of TESE is when tubules with mature spermatozoa (Johnsen score ≥8) are found in the histopathology specimen, irrespective of the overall state of spermatogenesis. The findings suggest that a lower limit threshold value of 2% of tubules with spermatogenesis in the histopathology specimen will result in a positive sperm retrieval.
  • However, it is not practical to perform a diagnostic biopsy before TESE because this would mean that patients undergo two surgeries, which adds to the cost and increases the complications.
  • The diagnostic biopsy is best coupled with an initial TESE before starting the ICSI cycle. Based on the findings of the histopathology specimen, patients may be then offered a repeat TESE if more sperm is needed on the day of ovum pick-up and ICSI. Also, if the initial TESE was negative, the biopsy result will help in the decision to offer a repeat TESE. This regimen is more cost-effective because the ICSI cycle will be started only if adequate sperm is retrieved.

Objective

  • To assess whether testicular histopathology can predict the outcome of testicular sperm extraction (TESE) in men with non-obstructive azoospermia (NOA) and therefore the role of preoperative diagnostic testis biopsy.

Patients and Methods

  • The study comprised a retrospective analysis of 388 patients with azoospermia who were referred from 2005 to 2010.
  • Information collected included a clinical history and an examination including age and testicular size, serum follicle-stimulating hormone, two semen analyses and testicular histology collected at the time of surgical sperm retrieval (TESE or microdissection TESE).

Results

  • In total, 388 patients with a mean (range) age of 37 (18–66) years were included in the present study.
  • Based on the history, clinical and laboratory findings, 112 patients had obstructive azoospermia and 276 patients had NOA.
  • All patients in the obstructed group had a positive sperm retrieval. The sperm retrieval rate for the NOA group was 50%.
  • An analysis of the results showed that the best predictor of a positive sperm retrieval was when tubules with mature spermatozoa were seen at biopsy, irrespective of the overall state of spermatogenesis (P < 0.001).

Conclusions

  • The presence of tubules with spermatazoa on biospy is the best predictor of a positive surgical sperm retrieval in patients with NOA.
  • The diagnostic biopsy is best coupled with an initial TESE before starting the intracytoplasmic sperm injection (ICSI) cycle.
  • Based on the findings of the histopathology specimen, patients may be offered a repeat TESE if more sperm is needed on the day of ovum pick-up and ICSI, or a redo TESE if the initial TESE was negative.
Abbreviations
ICSI

intracytoplasmic sperm injection

NOA

non-obstructive azoospermia

OA

obstructive azoospermia

SCO

Sertoli cell only

TESE

testicular sperm extraction

Introduction

Azoospermia occurs in 1% of men and in 10–15% of the infertile male population. Non-obstructive azoospermia (NOA) is caused by testicular failure, which may be primary (testicular) or secondary to hypothalamo-pituitary dysfunction (pre-testicular), and represents 60% of all cases of azoospermia. Obstructive azoospermia (OA) is caused by a blockage in the genital tract or ejaculatory dysfunction (post-testicular) [1, 2].

The management of patients with NOA relies on testicular sperm extraction (TESE) (or microdissection TESE) followed by intracytoplasmic sperm injection (ICSI) [3].

The recovery of spermatozoa is successful in only 50% of cases and therefore the ability to predict those patients with a high probability of achieving a successful sperm retrieval would be of great value in counselling the patient and his partner [4-6]. There is no single clinical finding or investigation that can accurately predict the outcome of TESE [7, 8].

In general, the smaller the testicular volume, the worse the prognosis, although there is no lower limit of testicular volume that excludes the presence of spermatozoa [7].

Similarly, studies have reported that higher serum FSH concentrations are inversely related with the probability of the presence of sperm in the testes, although the serum FSH concentration has a poor predictive value for a successful TESE. This is a result of FSH concentrations being inversely related to the total number of testicular germ cells present and not to the more advanced stages of spermatogenesis [9, 10].

Serum inhibin is also an inadequate predictor of successful TESE, although it is slightly more sensitive than FSH [11].

Y chromosome microdeletions are present in up to 8% of the infertile male population, with AZFc microdeletions being the most common. Although sperm recovery is possible in patients with AZFc microdeletions, spermatogenesis is absent in patients with complete deletions in the AZFa and/or AZFb regions. Thus, a Y chromosome microdeletion test will be of negative predictive value only in patients with AZFa and/or AZFb microdeletions [12-14].

Other suggested predictors of a positive TESE include the presence of spermatids in the ejaculate and the use of scrotal duplex profusion, although these are not widely accepted [15, 16].

Previous studies [10, 17-20] have concluded that testicular histology is the best predictor of a successful or unsuccessful TESE. The present study aimed to determine whether testicular histopathology can predict the TESE outcome for NOA patients, as well its value in the management of these patients.

Patients and Methods

The present study is a retrospective analysis of 388 patients with azoospermia who were referred to the uro-andrology clinic between 2005 and 2010. All patients were confirmed to be azoospermic using at least two semen analyses. A relevant clinical history was recorded, including age, history of an undescended testis, mumps orchitis, previous genito-urinary infection, radiotherapy, chemotherapy, surgical procedures or exposure to gonadotoxins. A clinical examination included secondary sexual characteristics, testicular size and consistency, epididymal distension, the presence of the vasa and a varicocele. All patients had their serum FSH concentrations measured, with a value in the range 1.5–12.4 IU/L being taken as normal.

Under general anaesthetic, all patients had surgical sperm retrieval in the form of TESE or micro-TESE. In both surgical procedures, a median raphe incision was made in the scrotum, the tunica vaginalis opened and the testis delivered. In TESE, four random, transverse, 1-cm incisions were made in each quadrant of the testis and a large piece of testicular tissue was taken from each. In micro-TESE, an equatorial incision involving three-quarters of the circumference was made using the surgical microscope to avoid vascular injury. Microdissection was then performed to expose the seminiferous tubules and multiple tiny pieces of testicular tissue were taken from areas that had dilated tubules. In both surgical procedures (TESE or micro-TESE), the specimens were examined by an embryologist where all the tubules were teased and analyzed for the presence of sperm. Any viable sperm present was cryopreserved for future use in ICSI. During the same surgical procedure, a histopathology specimen was placed in Bouin's solution and sent for analysis.

Because micro-TESE is a more costly and technically demanding procedure, requiring a skilled microsurgeon, it was reserved for NOA patients with very small testicular size (≤3 cm testicular length) to minimize the need for testosterone replacement therapy.

Based on the most predominant histopathological pattern, testicular histology was classified into: normal spermatogenesis; hypospermatogenesis (i.e. a reduction in the number of normal spermatogenetic cells); maturation arrest (i.e. an absence of the later stages of spermatogenesis); and Sertoli cell only (SCO) (i.e. the absence of germ cells in the seminiferous tubules).

In addition, a Johnsen score was determined for each testicular biopsy. In total, 100 tubules were examined per slide and each slide was scored on a scale of 1–10 based on the level of spermatogenesis (Table 1) [21].

Table 1. Johnsen score [21]
ScoreLevel of spermatogenesis
10Full spermatogenesis
 9Slightly impaired spermatogenesis
 8Less than five spermatozoa per tubule
 7No late spermatids; many early spermatids
 6Few early spermatids; arrest of spermatogenesis at the spermatid stage
 5Many spermatocytes
 4Few spermatocytes; arrest of spermatogenesis at the primary spermatocyte stage
 3Spermatogonia only
 2No germ cells; Sertoli cells only
 1No seminiferous epithelial cells; tubular sclerosis

The mean Johnsen score, highest Johnsen score and percentage of tubules with a Johnsen score ≥8 were calculated for each testicular biopsy.

Out of the 388 patients included in the present study, 338 had multiquadrant TESE and 50 had microdissection TESE. Sperm retrieval (TESE or micro-TESE) was performed bilaterally in 335 patients with a diagnostic biopsy taken from each testis (670 testes). The remaining 53 patients had unilateral sperm retrieval (TESE or micro-TESE) with a diagnostic biopsy taken from that side only (53 testes). Thus, 723 testes were investigated for histopathology and TESE outcome.

Unilateral sperm retrieval was performed in 53 patients because the contralateral testis had been removed for undescended testis (n = 15), testicular tumour (n = 6), torsion (n = 5), testicular atrophy (n = 11) and obstructive aetiologies in 16 patients.

All statistical analyses were performed using SPSS, version 16.0 (SPSS Inc., Chicago, IL, USA).

Results

In total, 388 patients with a mean (range) age of 37 (18–66) years were included in the present study. Based on history, clinical and laboratory findings, 112 patients had OA and 276 patients had NOA. The different aetiologies for each group are listed (Tables 2, 3).

Table 2. Aetiology of obstructive azoospermia (OA)
Aetiology (OA group)Patients, n (%) (N = 112)
Idiopathic25 (22.4)
Vasectomy24 (21.3)
Anejaculation19 (17)
Epididymitis18 (16)
Vasal aplasia12 (11)
Surgery5 (4.2)
Ejaculatory duct obstruction3 (2.7)
Absent left testis and right epididymal obstruction1 (0.9)
Blind ending vas1 (0.9)
Left absent vas and right ejaculatory duct obstruction1 (0.9)
Young's syndrome1 (0.9)
Congenital fusion of both testes1 (0.9)
Testicular tumour1 (0.9)
Table 3. Aetiology of non-obstructive azoospermia (NOA)
Aetiology (NOA group)Patients, n (%) (N = 276)
Idiopathic147 (53.2)
Bilateral undescended testes23 (8.3)
Unilateral undescended testis23 (8.3)
Mumps21 (7.6)
Varicocele (contributing factor)16 (5.8)
Testicular tumour8 (2.9)
Hypogonadotopic hypogonadism7 (2.5)
Testicular torsion5 (1.8)
Klinfefelter's syndrome5 (1.8)
Chemotherapy3 (1.1)
Y chromosome microdeletion3 (1.1)
Epididymo-orchitis3 (1.1)
Mixed gonadal dysgenesis3 (1.1)
Spinal cord injury2 (0.7)
Testicular trauma2 (0.7)
Empty sella syndrome1 (0.4)
Kallman syndrome1 (0.4)
Leprosy1 (0.4)
Liver disease1 (0.4)
Peno-scrotal lymphodaema1 (0.4)

OA

In the OA group (n = 112), the mean (range) patient age was 37 (18–66) years. A bilateral TESE was performed in 97 patients and a unilateral TESE was performed in 15 patients, giving a total of 209 testicles available for evaluation. All patients had a normal FSH concentration, with a mean (SD) testicular length of 5 (0.4) cm. The testicular histopathology in this group showed normal spermatogenesis in 105 patients, hypospermatogenesis in five patients and maturation arrest unilaterally in two patients.

All patients had some tubules with spermatozoa within, as shown by the mean (SD; range) percentage of tubules with a Johnsen score ≥8 of 84.4% (16.2%; 7–100%). The overall mean (SD; range) Johnsen score was 8.2 (1; 3.7–10).

As expected, all patients in this group (n = 112) had a positive sperm retrieval.

NOA

In the NOA group (n = 276), the mean (range) patient age was 36 (18–65) years. Male age was not associated with a positive sperm retrieval (P = 0.06).

A bilateral TESE was performed in 238 patients and a unilateral TESE was performed in 38 patients, giving a total of 514 testicles available for evaluation.

Out of the 514 testes sampled in the NOA group, 219 (42.6%) were positive for sperm, equating to 50% of patients with NOA storing sperm for ICSI.

The mean (SD; range) testicular length was 3.6 (0.9; 1–5) cm. Increasing testicular size was highly significant in predicting a positive sperm retrieval (P < 0.001) (Fig. 1). The FSH concentration was normal in 120 (43%) patients, inceased less than twofold in 88 (32%) patients, increased more than twofold in 66 (24%) patients and low in two (1%) patients. There was a strong negative correlation between serum FSH concentrations and the sperm retrieval rates (r = −0.208, P = 0.001) (Fig. 2).

Figure 1.

Relationship between testicular length and sperm retrieval.

Figure 2.

Relationship between serum FSH concentration and sperm retrieval.

The overall histopathological diagnosis for these 514 testes showed normal spermatogenesis, hypospermatogenesis, maturation arrest and SCO in 5%, 16%, 29% and 50% of the examined testes, respectively. There was a statistically significant decrease in the sperm retrieval rates with the decline in testicular histopathology from each subgroup; normal spermatogenesis to hypospermatogenesis to maturation arrest to SCO (P < 0.001) (Table 4).

Table 4. Sperm retrieval vs testicular histology in non-obstructive azoospermia
HistologyTestes, n (%) (N = 514)Testes with positive TESE, n (%) (N = 219; 42.6%)
  1. TESE, testicular sperm extraction.
Normal Spermatogenesis26 (5)26 (100)
Hypospermatogenesis82 (16)77 (94)
Spermatogenic arrest151 (29)56 (37)
Sertoli cell only255 (50)60 (24)

The mean (SD; range) Johnson score was 3.9 (2.4; 1–9.2), the mean (SD; range) for the highest Johnson score was 5.7 (3.2; 1–10) and the mean (SD; range) for the percentage of tubules with a Johnsen score ≥8 was 18% (30%; 0–100%).

There was a strong positive correlation between the presence of tubules with spermatozoa present (Johnsen score ≥8) in the testicular histopathology specimen and the sperm retrieval rates (r = 0.686, P < 0.001) (Fig. 3).

Figure 3.

Percentage of tubules with a Johnsen score ≥8 in the testes with a positive sperm retrieval (PSSR) outcome.

The mean for the percentage of tubules with a Johnsen score ≥8 for the testes with a positive sperm retrieval (n = 219) and negative sperm retrieval (n = 295) was 39.7% and 2%, respectively.

A discrepancy between the right and left testes regarding TESE outcome was seen in 30 patients with NOA. In 25 of these patients, the side from which sperm was retrieved had a percentage of tubules with spermatozoa present (Johnsen score ≥8) in the biopsy specimen, whereas, in the other five patients, there were no spermatozoa present in the biopsy specimen on both sides.

Analysis of the results showed that, out of age, testicular size, serum FSH concentration, overall histopathological diagnosis, mean Johnsen score and highest Johnsen score, the presence of tubules with a Johnsen score ≥8 in the histopathology specimen is the best predictor of a positive sperm retrieval (Fig. 4).

Figure 4.

Receiver-operating characteristic (ROC) curves showing that the presence of tubules with Johnsen score ≥8 is the best predictor of positive sperm retrieval.

Discussion

The use of surgically-retrieved sperm from the testis or epididymis with ICSI has now made it possible for patients with NOA and patients with OA not amenable to surgical reconstruction to father children. Sperm retrieval rates in OA are ≈100% [22, 23]. This was mirrored in the present study where all 112 patients had sperm retrieved despite two testes with maturation arrest. All the testes in this group had tubules with mature spermatozoa (Johnsen score ≥8 with a lower limit of 7%) in the histopathology specimen.

However, in NOA, a sperm retrieval rate of 50% can be expected, as reported in the present study [4-6].

At the initial consultation, couples often want to know the chances of sperm being retrieved. A failed TESE can have significant emotional and financial implications on the couple and this may be reduced by having preoperative knowledge of the likelihood of success [17, 24].

As yet, there are no absolute defined prognostic parameters for spermatogenesis being present in patients with NOA. Data such as testicular volume, serum FSH concentrations and the presence of associated male pathologies cannot be used as predictive factors of success [3, 19].

Several studies have concluded that a diagnostic testicular biopsy is the strongest indicator for sperm retrieval [22, 23, 25] and this is supported in the present study.

The results have shown that out of patient age, testicular size, serum FSH concentration, overall histopathological diagnosis (normal spermatogenesis, hypospermatogenesis, spermatogenic arrest, SCO), mean Johnsen score and highest Johnsen score, the strongest predictor of success is when tubules with mature spermatozoa (Johnsen score ≥8) are found, irrespective of the overall state of spermatogenesis. The findings suggest that a lower threshold limit of 2% of tubules with spermatogenesis in the histopathology specimen will result in a positive sperm retrieval.

This has been confirmed in other studies reporting that the presence of elongated spermatids or spermatozoa in a diagnostic biopsy is related to a high sperm retrieval rate for ICSI [17, 18, 26-29].

Testicular volume and plasma FSH concentrations are good clinical parameters for the prediction of sperm retrieval but less accurate than testicular histopathology. This is supported by the fact that many patients with maturation arrest are found to have normal plasma FSH concentrations and testicular volume [30]. Variations in plasma FSH concentration can arise for reasons unrelated to spermatogenesis; similarly, testicular volume shows a wide variation as a result of many factors, including racial variation [31]. Moreover, the epididymis or scrotum can obscure the true measurement of the testicular volume when performed clinically [32].

Although the results of the present study have not shown any statistically significant correlation between patient age and a positive sperm retrieval, the relationship appeared to be in the positive direction; however, there was lack of diversity with respect to age groups. Thus, studies with larger sample sizes and more diverse age groups are warranted to enable a more adequate assessment of the effect of age on sperm retrieval.

Based on the fact that testicular histopathology is the most accurate predictor of a positive TESE, some studies have suggested performing a diagnostic biopsy before TESE [18, 28]. However, any advantage to carrying out such a procedure is debatable because the diagnostic biopsy itself is an invasive procedure that may have complications similar to the TESE procedure (e.g. infection, bleeding and haematoma, tubular sclerosis or a permanent devascularization of the testes) [24]. This would mean that patients undergo surgery twice, which adds to the cost and increases the chance of complications. Moreover, because testicular tubules of patients with NOA are usually heterogeneous, the absence of sperm in a single biopsy does not guarantee a similar pattern in the entire testis [7].

Studies are available suggesting that there is no significant difference in live birth rates between fresh and frozen sperm in ICSI for men with NOA; it would appear preferable to perform any diagnostic biopsy at the same time as the initial TESE in isolation, before commencing any ICSI cycle [33-36].

On the day of oocyte collection, a second TESE could be performed in patients with an initial positive TESE only if an inadequate number of sperm is found after thawing. This is usually predicted by the embryologist based on the amount of sperm retrieved in the initial TESE and the subsequent histopathology.

As for patients in whom the initial TESE was negative, a second redo TESE is warranted only in the presence of favourable histopathology in the diagnostic biopsy (i.e. the presence of at least 2% of the tubules with a Johnson score ≥8).

Repeat and redo TESE are best performed microsurgically (microdissection TESE), after a 6-month recovery period, and by an experienced andrologist, aiming to increase the chances of sperm retrieval and minimize the complications [24, 37].

Patients with a negative TESE and unfavourable histopathology should be informed that the chances of sperm retrieval from a second TESE will be extremely low.

This regimen is more cost-effective because the ICSI cycle will be started only if adequate sperm is retrieved. Couples will be able to know before starting the cycle if sperm is present or not. In the case of a negative first TESE, couples may be counselled about the feasibility of attempting another TESE based on the results of the histopathology. Couples with a negative TESE and unfavourable histopathology will have time to consider other options.

Some patients may already have had a testis biopsy performed in isolation and, in this situation, providing that the pathology was favourable, they could go on to have a fresh TESE and ICSI cycle with the knowledge that sperm will be retrievable on the day.

Another important benefit of histopathology is the detection of carcinoma in situ of the testis, which occurs in 1–3% of patients with severe male factor infertility, and which progresses to testicular cancer. Both carcinoma in situ and seminoma of the testis have been reported in patients undergoing TESE [38, 39].

A histopathology showing normal spermatogenesis will also accurately diagnose cases of OA that were missed clinically. Such cases may be offered subsequent surgical exploration with possible reconstruction. However, up to 15% of patients with OA have an intratesticular obstruction that is not reconstructable [40]. It should be noted that patients with OA may have some impairment in spermatogenesis. In two patients with obstructive azoospermia secondary to anejaculation and epididymitis, respectively, the histopathology showed maturation arrest on one side. This may be a result of the longstanding obstruction causing damage to the spermatogenic tubules. It should also noted that, although patients with NOA have impairment of spermatogenesis, it may be possible for these patients to have areas in the testes showing foci of normal spermatogenesis. In the present study, 26 testes sampled from 18 patients had the histopathological diagnosis of normal spermatogenesis. These patients were still considered to have NOA rather than intratesticular obstruction. In those patients, the aetiology of azoospermia, the testicular size, serum FSH concentrations and the intra-operative findings were indicative of NOA rather than OA.

A cause for azoospermia was not identified in 22% and 53% of patients with OA and NOA, respectively. Those patients probably have an unidentified genetic basis for infertility (Tables 2, 3). As a result of additional research into the genetic basis of male infertility, similar cases could be treatable with gene therapy.

In conclusion, testicular histopathology gives valuable information with regard to the likelihood of retrieving sperm at TESE. This is best performed at the time of TESE, aiming to negate the need for a second operation. It is useful for determining whether a negative TESE should be repeated, to confirm any diagnosis, particularly malignancy, and to allow patients to understand the results of the assisted conception process.

Conflict of Interest

None declared.

Ancillary