Clinical Conditions


Azoospermia due to low sperm production (non-obstructive azoospermia) affects approximately 1% of the male population and 10% of men who seek fertility evaluation. Testis biopsy reveals that these men have Sertoli cell-only pattern, maturation arrest, or hypospermatogenesis. Until recently, it was assumed that men with non-obstructive azoospermia were untreatable. Indeed, these patients were often referred to as being "sterile" or having "testicular failure." The only way these couples could have children was to use donor spermatozoa or to adopt. Several observations have changed our approach to this condition. First, we have observed that direct evaluation of testis biopsy specimens often demonstrates sperm in men with non-obstructive azoospermia, despite severe defects in spermatogenesis (Jow et al., 1993). This observation has been interpreted to mean that a low level of sperm production may be present in the testes of men with azoospermia, but the sperm do not survive epididymal transit and ejaculation (Silber et al., 1997).

In addition, it was previously thought that sperm must traverse the male reproductive tract before acquiring the ability to normally fertilize an egg. The success of treatment of men with obstructive azoospermia using sperm extracted from the epididymis or testis has changed this view. Although testicular sperm have dramatically lower motility than those that have transited the male reproductive tract, these sperm can be used for intracytoplasmic sperm injection (ICSI) during in vitro fertilization (IVF). Such observations led investigators to perform testicular sperm extraction (TESE) with ICSI for men with non-obstructive azoospermia. Low pregnancy rates of 20 to 21% per attempt have been reported. The following is our approach to evaluation and attempted sperm retrieval for men with non-obstructive azoospermia.

Patient Evaluation

To determine if a semen sample is truly azoospermic, centrifugation of the semen sample with meticulous microscopic examination of the pellet is necessary. Although this might seem obvious, Ron-El et al. (1997) reported that sperm was found on extended sperm analysis of a centrifuged semen specimen in up to 35% of men who were thought to have non-obstructive azoospermia. In addition, we have found that up to 10-20% of men who have inadequate sperm on preoperative semen analysis will actually have sperm usable for ICSI on the day of oocyte retrieval. Therefore, we always repeat a semen analysis on the day of planned sperm retrieval for men with non-obstructive azoospermia.

For all patients with azoospermia, a complete history and physical examination is necessary to identify potentially correctable causes of male factor infertility. Typically, the man with non-obstructive azoospermia will have small testes (< 15 cc) with a flat epididymis. Some men may have a history of cryptorchidism. Hormonal evaluation of a man with non-obstructive azoospermia (NOA) will typically demonstrate an elevated serum FSH, with normal or nearly normal testosterone and estradiol levels. Prior to further intervention, we will usually treat any correctable abnormalities that are found on evaluation of a man with NOA, including surgical repair of large varicoceles, correction of hormonal abnormalities, and avoidance of gonadal toxins for at least three months prior to attempted TESE. Orchiopexy and varicocele repair are usually only considered if female age is less than 38, because with advanced female age the window of opportunity to achieve pregnancy is limited and these procedures benefit only a small percentage of men and require six months of recovery prior to attempted sperm retrieval.

However, with advanced female age (>38) the window of opportunity to achieve a pregnancy is limited, making orchiopexy and varicocele repair not worthwhile because they benefit only a small percent of men and require six months of recovery prior to attempted TESE.

Scrotal ultrasound

The ASRM/AUA Practice Guidelines do not recommend routine application of scrotal ultrasound for evaluation of the infertile male, although ESHRE guidelines do support routine ultrasound examination. Scrotal ultrasound should be strongly considered for patients with a history of cryptorchidism or prior germ cell tumors, and for men who have any unexplained abnormalities on testicular examination. In addition, it should be considered f body habitus limits physical examination of the scrotum or if the findings on physical examination are equivocal. Routine use of a scrotal ultrasound to screen for subclinical varicoceles is not recommended, as repair of very small varicoceles has not been shown to be effective in improving semen parameters or fertility. Ultrasound may also be of value in following patients after testicular biopsy, since intratesticular hematomas or scar tissue formation commonly occur and affect the timing for subsequent sperm retrieval procedures.

Genetic Abnormalities and Testing

In this section, I will emphasize common genetic disorders associated with spermatogenic failure, including non-obstructive azoospermia. These abnormalities include both chromosomal abnormalities, detectable with routine karyotype testing, and Y chromosome microdeletions, so called "AZF defects." Other rare genetic causes of male infertility are nicely reviewed in Mak and Jarvi: J Urology 156:1245-57, 1996. For men with severe male factor infertility, including sperm concentrations less than 10 x 106/cc and non-obstructive azoospermia, karyotype evaluation and Y chromosome microdeletion analysis is recommended before treatment with assisted reproduction.

Evaluation of a sequential series of 170 men with non-obstructive azoospermia who were candidates for TESE at Weill Cornell revealed that 17% of these men had definable genetic defects (Y chromosome microdeletions or karyotype abnormalities; Rucker et al., 1998.) We have found that the knowledge of having a genetic defect leads many men to pursue options other than TESE-ICSI and that, regardless of treatment choice, the majority of men find it reassuring to know the cause of their infertility.

Karyotype evaluation

The most common karyotypic abnormality in men with severe male factor infertility is Klinefelter syndrome, affecting up to 7-13% of azoospermic men. Almost all men with the "pure, classic form" (47,XXY) of Klinefelter will be azoospermic, whereas limited sperm production is commonly found in men with a mosaic pattern of Klinefelter syndrome. It was previously felt that only spermatogonia with a 46,XY complement could produce spermatozoa (Martini et al., Hum Reprod 11:1638, 1996); however, recent observations indicate that a significant proportion of 24,XY spermatozoa are present in the testes of men with Klinefelter (Cozzi et al., Hum Genet 93:32, 1994). General teaching has suggested that men with Klinefelter syndrome can be readily identified by their typical physical appearance of tall stature, gynecomastia and small, firm testes. Again, however, recent observations contradict this statement. I have detected two men in my practice who were normally masculinized, between 5'6" and 5'10" in height, but had non-mosaic Klinefelter syndrome. Observations reported by Oates et al. (J Urol 155:476A, abstract 660; 1996) at the AUA Annual meeting also confirm that some men with chromosomal abnormalities will have an otherwise normal phenotypic appearance except for their infertility. Most men with Klinefelter syndrome have sperm retrievable with testicular sperm extraction (TESE) and can have children with ICSI. We have attempted treatment of over 15 men with this condition and eight children have been born to these couples. Other karyotypic abnormalities identified include Robertsonian translocations, chromosomal inversions and sex chromosome abnormalities.

Y chromosome (AZF) microdeletions

Several genes have been identified in the distal portion of the long arm of Y (Yq) that are frequently deleted in men with non-obstructive azoospermia. The best described gene has multiple copies, and is referred to as DAZ (deleted in azoospermia.) Deletions involving DAZ were identified in 13% (12/89) of azoospermic men screened by Reijo et al. in 1995. In addition, Reijo et al's evaluation of men with severe azoospermia revealed that 6% (2/35) had DAZ deletions. Other investigators have found longer deletions of the Y chromosome associated with male infertility. Vogt et al. have suggested that three relatively discrete regions of Yq, AZFa, AZFb, and AZFc, are deleted in severely infertile men and that the deleted region determines the chance of sperm production. Reijo et al. have shown similar deletions in men with different levels of sperm production, suggesting that other genetic or environmental factors may affect the phenotype of sperm production in men with Y chromosome microdeletions.

Several investigators have found that 3-18% of men with severe sperm production abnormalities, including azoospermia, have Y chromosome deletions. However the literature is difficult to evaluate because the data was generated in multiple laboratories, each looking at different patient populations and examining different regions and sites on Yq. For example, some investigators considered a microdeletion present when only a single ?(STS) was absent, while others considered a microdeletion present only if sequential sites on the Y chromosome failed to amplify with PCR-based analysis. Nevertheless, all investigators consistently found Y microdeletions in a measurable proportion of severely subfertile men, with no detectable deletions in normal fertile men, nor in the fathers or brothers of men with Y microdeletions.

Y chromosome deletions affecting fertility usually involve deletion of one or more entire AZFa, AZFb, or AZFc regions. An additional region of the Y chromosome referred to as AZFd has been described, however, it has been determined that AZFd is within AZFc, has no prognostic significance, is not associated with impaired sperm production, and therefore such deletions should not be reported, as they are clinically irrelevant. The specific region that is missing on the Y chromosome may provide prognostic significance. Approximately 50% of men with deletions involving only AZFc, have sperm present in the ejaculate. In azoospermic men with AZFc deletions, sperm production is commonly present within the testicle, and TESE is as successful as for other men with non-obstructive azoospermia. At Weill Cornell, sperm was found by TESE in 70% of men with AZFc deletions and azoospermia..

For men with deletions involving the AZFb region, the chance of having sperm in the ejaculate or finding sperm with TESE is severely decreased. Sperm was found in zero out of 23 men we evaluated with deletions involving AZFb who had a biopsy or sperm retrieval attempted with TESE (Hopps et al., 2003). Silber et al. also reported that in a series of five men with deletions involving both AZFb and AZFc no sperm was found on attempted sperm retrieval (ASRM Annual meeting, P-184, 1999). Therefore, we do not recommend that men with deletions involving the entire AZFb region undergo TESE unless use of donor sperm is planned as a back-up.

Deletions involving the entire AZFa region are also commonly associated with a Sertoli cell-only pattern on diagnostic biopsy (Kamp et al., 2001). Overall, approximately 9% of men with Sertoli cell-only pattern are estimated to have AZFa deletions. The need to discriminate between partial and complete deletions of an AZF region is reflected in the observation that at least one patient with a deletion of part of the AZFa region had germ cells on testis biopsy, however, to-date no sperm has been retrieved from men with complete deletions of AZFa or AZFb. Because the documented number of cases in the literature is limited, absolute predictive statements are not possible to make at this time. However, the prognosis is clearly different and dramatically worse for men with complete AZFa and AZFb deletions than for other patients with non-obstructive azoospermia.

Because Y chromosome abnormalities, including deletions, will be passed on to any male child who is produced after assisted reproduction, these men must have genetic counseling prior to treatment. In an important preliminary study of fathers and ICSI-derived sons, Kent-First et al. (Lancet 348:332, 1996) found that 10% (3/32) of unselected ICSI-derived boys had detectable Y chromosome microdeletions, however, only one of the three boys had a father with Y chromosome microdeletions detected on testing of his peripheral blood. These results suggest that mosaicism with germ-line deletions on the short segments of the Y chromosome frequently develop in spermatozoa of men with severe male factor infertility. Since men with these genetic defects have rarely or never fathered children naturally, it is uncertain whether any medical conditions will be present in the offspring with Y chromosome microdeletions, except for infertility. This knowledge makes genetic counseling difficult. On the other hand, common sense suggests that because the fathers are otherwise healthy and normal, the presence of a Y chromosome microdeletion does not pose a high risk for major congenital defects in potential offspring. However, the answer will not be available for many years, when the children born from this process are adults.

Hormonal therapy

Hormonal therapy has never been demonstrated in randomized controlled trials to increase sperm production or quality in men with normal hormone levels. Therefore, I use hormone therapy only in men with demonstrated hormone abnormalities. Hormonal evaluation of a man with non-obstructive azoospermia (NOA) will typically demonstrate an elevated serum FSH, with normal or nearly normal testosterone levels. However, many men with NOA have an abnormal testosterone (ng/dL) to estradiol(pg/mL) ratio (T/E2 ratio), which is correctable with treatment using aromatase inhibitors like testolactone or anastrozole (Pavlovich et al., 2001). Whereas normal fertile men will have a T/E2 ratio of 16 + 3; men with NOA have a ratio of 7, and men with Klinefelter syndrome have a ratio of 4. The inbalance is correctable with aromatase inhibition, or delivery of exogenous testosterone, suggesting that increased aromatase activity from relative Leydig cell hyperplasia within the testicle may be the cause of this phenomenon. Correction of abnormal T/E2 ratios in men with severe oligospermia can effect dramatic improvements in sperm concentration and motility. We now routinely evaluate testosterone and estradiol levels for men with non-obstructive azoospermia or severe oligospermia. Men with low testosterone and a low T/E2 ratio are routinely treated with anastrazole, one miligram per day (Raman & Schlegel, 2002.) For men with Klinefelter syndrome, response is better using testolactone, 50-100 milligrams per day. Testosterone levels increased from 190 ng/dL to 332 ng/dL during testolactone treatment of these Klinefelter syndrome patients, with improved T/E ratios.

Varicocele repair

The role of the varicocele in male infertility continues to be controversial. As early as 1987, the World Health Organization stated that varicoceles are associated with decreased male fertility. Numerous uncontrolled studies have suggested increased sperm count, motility and morphology after varicocele repair. However, few randomized controlled studies have documented a beneficial effect of varicocele repair on male fertility. A recent meta-analysis published in Lancet reported little beneficial effect of varicocele repair on male fertility (Evers & Collins, 2003). Unfortunately, this review included studies in which patients were treated for "subclinical varicoceles," a practice that is not condoned by even the most ardent supporters of varicocele repair.

Of more relevant interest is the increasing practice of varicocele repair for men with non-obstructive azoospermia. One recent study found a return of sperm to the ejaculate in 43% (12/28) of men with non-obstructive azoospermia at an average follow-up of 24 months. A second study found spermatogenesis adequate to produce sperm in the ejaculate of 55% (12/22) of men with non-obstructive azoospermia. Other smaller and older studies have conflicting results. I recently reviewed a series of patients with varicoceles and non-obstructive azoospermia who were evaluated and treated at Weill Cornell. Of 31 men who underwent varicocele repair for documented non-obstructive azoospermia, 20% (7/31) had sperm on at least one postoperative semen analysis. However, only 9.6% (3/31) of men after varicocele repair had adequate motile sperm in the ejaculate for ICSI (without TESE). A history of prior varicocele repair did not affect the results of TESE, for men with non-obstructive azoospermia and varicoceles. Retrospective analysis of patients with clinical varicoceles identified before TESE shows that the rate of sperm retrieval was identical in those who had their varicoceles repaired before TESE as compared with those who did not have them repaired before TESE, 60% (41/68) and 60% (42/70) respectively (Schlegel & Kaufman, 2004.) Based on these data, varicocele repair is of limited value for men with non-obstructive azoospermia and varicoceles. In my opinion, surgical repair should probably be reserved for men who are very young and have large varicoceles, and possibly for those with testicular atrophy associated with a varicocele.

Diagnostic biopsies in men with presumed non-obstructive azoospermia

The diagnosis of NOA can only be definitively made on testicular biopsy. Testis biopsy can also rule out the unlikely possibility of testicular intratubular germ cell neoplasia (carcinoma-in-situ) that is more common in men with unexplained unilateral testicular atrophy or with a history of cryptorchidism.

The most advanced spermatogenic pattern, as opposed to the predominant pattern, appears to affect the results of sperm retrieval. For men who had at least one area of hypospermatogenesis present on diagnostic testis biopsy, spermatozoa were retrieved in 81% (31/39) of attempts, whereas for men with maturation arrest as the most advanced pattern, spermatozoa were retrieved inonly 42% (8/19) of attempts. If the entire diagnostic biopsy had a Sertoli cell-only pattern, then sperm were retrieved in 24% (5/21) of TESE attempts. Although no finding absolutely determined sperm retrieval or negated the possibility of successful TESE, the findings of diagnostic biopsy were helpful in evaluating the chance of success with TESE (Su et al., 1999). In addition to the role of diagnostic biopsy in identifying rare cases of intratubular germ cell neoplasia (carcinoma-in-situ) and confirming the diagnosis of non-obstructive azoospermia, diagnostic biopsy helps to predict the chance that a TESE procedure will obtain sperm.

When do you perform a diagnostic testis biopsy before TESE for ICSI?

Testicular biopsy will help to determine the prognosis for sperm retrieval, but it will not provide definitive proof of whether sperm will be found with a more intensive evaluation of the testis (TESE or microdissection TESE). Therefore, I do not require a diagnostic testis biopsy prior to TESE-ICSI for non-obstructive azoospermia. A diagnostic biopsy should be performed if the etiology of azoospermia is not clear, if the risk of carcinoma-in-situ is high (rare), or if the results of biopsy will affect the couple's choice to undergo TESE-ICSI. The reasons to determine the etiology are multiple (as previously discussed):

  1. Some men with obstructive azoospermia prefer reconstruction to sperm retrieval-ICSI.
  2. The genetic abnormalities associated with obstructive and non-obstructive azoospermia are different, and appropriate genetic testing cannot be done without knowing which defects to search for.
  3. There is a risk of not finding sperm in men with non-obstructive azoospermia. These couples should be counseled regarding the potential use of donor spermatozoa if sperm are not identified in the man. It is, likewise, inappropriate to discuss the use of donor sperm with men who have normal sperm production.
  4. The technique of sperm retrieval is different in obstructive & non-obstructive cases.

If a biopsy is performed, testicular tissue should be frozen for possible subsequent use. I perform diagnostic testis biopsies if a couple will not proceed to TESE-ICSI with only a 25% chance of sperm retrieval. If a uniform pattern of Sertoli cell-only is discovered, the couple has obtained useful information from the biopsy and avoided unnecessary ovarian stimulation.

Considerations Before Treatment

Sperm retrieval by testicular sperm extraction (TESE) can be performed prior to or coincident with an IVF cycle for the female partner. We can perform all initial TESE procedures on (or occasionally the day prior to) the day of oocyte (egg) retrieval during a programmed IVF cycle to maximize the potential to retrieve viable spermatozoa for use with ICSI. Therefore, the female partner must be evaluated by the IVF center before considering a TESE procedure.

Because men with NOA have marginal sperm production, TESE procedures should be delayed for at least 6 months after any intervention such as a prior biopsy or TESE procedure, or other inguinal/scrotal surgery (Schlegel & Su, 1997). Because the testicular blood supply penetrates the tunica albuginea and then disperses in a series of end-arteries that spread over the testicular parenchyma, multiple biopsies should be avoided to minimize the risk of devascularization of the testis. The use of optical magnification may also minimize the risk of testicular injury. We have seen two men who have had significant devascularization of the testis with atrophy after TESE performed at other centers using multiple biopsy technique (Schlegel & Su, 1997).

Fresh or Frozen?

Unfortunately, no preoperative parameter absolutely predicts the chance of sperm retrieval by TESE in men with non-obstructive azoospermia. Testicular volume and FSH levels have no predictive value, and as discussed above, even the histology of a diagnostic biopsy cannot absolutely predict sperm retrieval when the rest of the testis is sampled. Unfortunately, ovarian hyperstimulation is necessary for all women, even though there is, on average, only a 60% chance of sperm retrieval. Many couples choose to have donor sperm ready to use, in case sperm is not retrieved, to maximize each IVF cycle. . There are several good reasons to use simultaneous TESE with ICSI:

  1. Sperm retrieval is difficult in men with non-obstructive azoospermia (NOA).
    A single testicular biopsy is inadequate to retrieve sperm in most men with NOA. We found that less than one-half of successful sperm retrievals were accomplished on the first random biopsy. Careful dispersion of the testicular specimen and simultaneous evaluation by an experienced embryologist is needed to determine how many biopsies are needed. We have noted that sperm may not be found until the fourteenth random testicular biopsy in men with NOA (Ostad et al, 1998.) Microdissection TESE will increase the proportion of men with NOA who have sperm found over that achieved with random biopsies alone. A controlled comparison of microdissection versus random biopsies demonstrated that approximately one-third of men with sperm present in the testes have unsuccessful treatment with random biopsies alone. In addition, far less tissue is removed with microdissection, compared to random biopsies, despite having a higher yield of sperm (Schlegel, 1999.) Despite a single larger incision, less intratesticular reaction has been seen (on ultrasound) after microdissection when compared to a multiple biopsy approach. Therefore, it is now incorrect to consider TESE effectively performed with only a single biopsy per testes.
  2. Testis biopsy before the ICSI cycle is often an unnecessary operation.
    Many men who have had repeated azoospermic semen analyses will actually have enough sperm present in their ejaculate on the day of sperm retrieval to cancel a planned TESE procedure. For men at Weill Cornell, 20% of planned TESE-ICSI cases for "non-obstructive azoospermia" are cancelled because adequate numbers of viable sperm are found on the morning of the planned TESE procedure. Ron-El et al. have reported that up to 35% of men with NOA have sperm found on an extended sperm preparation, allowing men to avoid the unnecessary testicular procedure.
  3. Testicular sperm from men with NOA will often not survive freeze-thaw.
    Our experience is that testicular spermatozoa retrieved by biopsy from men with well-documented NOA will usually be non-viable after thawing. A total of 95 attempted ICSI cycles using frozen-thawed testicular sperm specimens were evaluated. All men with sperm retrieved in this series had non-obstructive azoospermia. Motile sperm were documented prior to cryopreservation. Out of the 95 cycles, only 33% (31/95) had viable sperm from the frozen-thawed sample. All other patients required repeat TESE. For the 31 cycles with frozen-thawed sperm alone, a clinical pregnancy rate of only 39% and a live birth rate of only 26% were observed. The clinical pregnancy rate with fresh sperm is 47%, and live birth rate is 31%. Not only do sperm frozen from men with non-obstructive azoospermia usually not survive freeze-thaw, but they appear to be less successful than freshly-retrieved sperm in effecting pregnancies. Another group from Germany reported that only 42% (178/426) of attempts at using frozen testicular tissue had enough motile sperm present after thawing to inject all oocytes. A total of 38% (161/426) of attempts had no motile sperm present for injection, and the remainder of the cycles required use of immotile spermatozoa because inadequate motile sperm were present in the thawed specimen to inject all oocytes (Fisher et al., [abstr O-190] ESHRE annual meeting, 1998).
  4. Testicular sperm from men with NOA have impaired function
    Fisher et al. further noted a dramatically lower pregnancy rate of 8% (+hCG/transfer) when immotile (frozen-thawed?) spermatozoa were used for injection for immotile sperm cycles, versus a pregnancy rate of 30% when motile frozen-thawed testicular spermatozoa where used. A higher spontaneous abortion rate after implantation was also seen for immotile sperm. Of additional concern, Ron-El et al. noted poor embryo cleavage after use of immotile spermatozoa, despite reasonable fertilization rates, suggesting that testicular sperm from men with NOA may have impaired function after freeze-thaw (abstract O-181, ESHRE annual meeting, 1998). These findings suggest that use of immotile frozen-thawed spermatozoa (the only sperm available for ICSI in most cases) causes poor embryo development with lower implantation and pregnancy rates, even if fertilization is effected. The poor results with these sperm suggest that the initial TESE cycle may be wasted (again, a potentially unnecessary operation) since most sperm don't survive freeze-thaw with motility.
  5. Subsequent ICSI cycles should await the potential for repeat TESE retrieval.
    Since poor results are obtained with non-motile frozen-thawed sperm, and most samples do not have enough motile spermatozoa to inject all oocytes, the back-up option of repeat TESE retrieval should be available to allow optimal results. Unfortunately, repeat sperm retrieval is not certain, and may require up to 6 months after the initial TESE to optimize the chance of obtaining sperm. In our experience, repeat TESE cycles retrieve sperm in only 25% of cases if performed within 6 months, versus 80% of cases if performed after 6 months of an initially successful TESE for NOA. Note, however, that the 80% success rate means that repeat TESE does not always yield sperm, even if they were obtained previously. (Schlegel & Su, 1997.)
    Although simultaneous sperm retrieval-ICSI carries the risk of ovarian hyperstimulation without the availability of testicular sperm, the above observations clearly suggest that the best results will be obtained using fresh testicular sperm for men with NOA. Unsuccessful retrieval attempts may be salvaged with donor sperm if couples choose to avail themselves of that option.

TESE (Testicular Sperm Extraction)

On the day of oocyte retrieval, scrotal exploration is performed through a median raphe incision under local or general anesthesia, and sperm are retrieved using an open testicular biopsy technique. In order to confirm accurate identification of the testis and to avoid any injury to the epididymis, delivery of the testis is routinely performed. Testicular blood vessels in the tunica albuginea are identified with 8-15x optical magnification. An avascular region near the midportion of the medial, lateral or anterior surface of the testis is chosen, and a generous incision in the tunica albuginea, avoiding any capsular testicular vessels, is created to directly examine a wide area of testicular parenchyma.

Microdissection TESE

The technique that we developed allows the removal of tiny volumes (2-3 mm?; 1-5 mL) of testicular tissue with improved sperm yield. This technique requires use of an operating microscope and there is a brief learning curve to identify which tubules contain spermatogenesis. The tubules containing sperm can often be visually identified under an operating microscope after opening the testis, when 15-25x magnification is used to assist the biopsies. This approach 1) improves the yield of spermatozoa per biopsy, 2) results in less tissue removal (and loss of testicle), 3) makes the embryologist's job easier in finding sperm, since less tissue has to be examined and 4) allows identification of blood vessels within the testicle, minimizing the risk of vascular injury and loss of other areas of the testis (Schlegel, 1999). Our observations of better sperm yield when using microdissection TESE (versus multiple random biopsy TESE) has been confirmed by several other investigators (Amer et al., 2000; Okada et al., 2002; Okubo et al., 2002).

The excised testicular biopsy specimen is placed in human tubal fluid culture medium supplemented with 6% Plasmanate. With this approach, sperm yield can be enhanced 2-fold per biopsy specimen (from 63,822 to 164,326 sperm/specimen), despite excision of 70-fold less testicular tissue (722mg versus 9.4mg). These result were demonstrated for a series of men with non-obstructive azoospermia who underwent standard testicular biopsies as controls in the same testes that were used for microdissection TESE. In a sequential series of patients who had sperm retrieval attempted with standard multi-biopsy TESE procedures vs. TESE with microdissection, the proportion of non-obstructive azoospermic men with sperm retrieved increased from 36% (4/11) to 68% (15/22) with the application of microdissection. For five men, 33% (5/15 men with sperm found with TESE), sperm could only be found in the testis with microdissection. Microdissection is of no therapeutic benefit for approximately 36% of attempted TESE cases, because no enlarged (normal) tubules are seen. Unfortunately, most of these cases also failed to yield spermatozoa with multiple standard biopsies. The use of microdissection, however, does still expedite sampling of testicular tissue, since multiple samples of seminiferous tubules can be rapidly examined and excised.

Handling of biopsy samples

When standard (larger volume) testicular biopsy samples are obtained, isolation of individual tubules from the mass of coiled testicular tissue is achieved by initial dispersal of the testis biopsy specimen between two sterile glass slides, which compress the testicular parenchyma to isolate individual seminiferous tubules. Subsequently, mechanical disruption of the tubules is accomplished by mincing the extended tubules with sterile scissors in HTF/Plasmanate medium (Schlegel et al., 1997). Additional dispersion of tubules is achieved by passing the suspension of testicular tissue through a 24-gauge angiocatheter (Ostad et al., 1998). For minimal tissue specimens, limited dissection is performed in the operating room, because the tissue sample is so small. Since spermatozoa are normally within the tubules, at least some of the tubules are cut and passed through a 24-gauge angiocatheter. These samples are then immediately examined in the operating room under phase contrast at 200x power to determine whether adequate numbers of sperm have been found for ICSI.

Dissection is completed when it has been reliably determined that enough sperm for ICSI are retrieved. If no spermatozoa are seen, then 1) additional biopsies of tissue are obtained through the same tunical incision, 2) biopsies are performed using additional incisions (rarely needed if the testis is initially opened widely), and 3) contralateral samples are obtained, if needed. Multiple biopsies carry benefit for the patient, as only 23% of men have sperm found on the first standard biopsy for non-obstructive azoospermia, and sperm may not be seen until the fourteenth biopsy (Ostad et al., 1998). Extended efforts to remove large numbers of sperm from the testis is of limited value, since the sperm may not survive freeze-thaw. Opening of the individual tubules may be done in the embryology laboratory, immediately prior to ICSI. After ICSI has been performed, excess aliquots of tissue are then processed for cryopreservation.


The general ICSI procedure has been described previously. Aggressive immobilization of spermatozoa, a technique that increases fertilization rates for immature sperm, is carried out for all testicular spermatozoa.


Results with ICSI are primarily dependent on the availability of viable sperm and age of the female member of the couple being treated. Encouraging experience has been obtained at Weill-Cornell with TESE-ICSI in the past 588 attempted treatment cycles for couples in whom men had nonobstructive azoospermia. The mean age of patients entering treatment was 36.1 for men and 32.1 years for women. In men, the initial mean serum FSH level was 22.4 IU/L (normal 1-9 IU/L), and average testis volume 9.6 cc. During the past 588 attempted TESE-ICSI cycles, sperm were retrieved for injection in 341 cycles (57.9% retrieval rate). For those cycles in which sperm were retrieved, the fertilization rate per injected oocyte was 56.0% (1994/3560), and embryo transfer occurred in 91.8% of cycles. Clinical pregnancies (fetal heartbeat on ultrasound) were established in 46.6% (159/341) of evaluable cycles and live deliveries occurred in 42% of cycles for which sufficient time had passed for completed gestation. A total of 198 children have been born from our center. Twin deliveries occurred in 11% of cases, triplets in <1%, and singleton deliveries for the remainder.

No etiology of azoospermia provided an absolute predictor for the presence or absence of sperm within the testes, except for AZFa and AZFb deletions. Testicular volume and serum FSH levels did not predict sperm retrieval. Thirty-seven men with classic, non-mosaic Klinefelter syndrome (47,XXY, or mosaic patterns that do not include 46,XY) provided sperm for 52 ICSI cycles at our institution, including three cycles in which cryopreserved sperm from previous TESE were used. Sperm were found in 69% (34/49) of the fresh retrieval cycles. Of the cycles in which sperm were retrieved by TESE, the fertilization rate per injected oocyte was 60%, embryo transfer occurred in 79%, clinical pregnancy was achieved in 47% (16/34),live delivery occurred in 43% (13/30), and there was a multiple gestation rate of 40%. Eighteen healthy children have been born (all 46,XX girls and 46,XY boys.) Pre-treatment testicular biopsy histology was not helpful in distinguishing who would succeed with microdissection TESE for patients with Klinefelter syndrome. Although 25 of these men had Sertoli cell-only on diagnostic biopsy, 76% of these patients had sperm found on subsequent microdissection TESE. Even though one of the patients treated had previously undergone multiple random biopsy TESE with no sperm found, sperm were retrieved in a subsequent procedure using the microdissection TESE technique.. These findings illustrate the potential for TESE-ICSI to provide fertility despite underlying genetic abnormalities.

Another treated subset of men with nonobstructive azoospermia includes the cohort of 34 men with a history of chemotherapy administered for a variety of diagnoses who underwent 41 sperm retrieval attempts with TESE for persistent nonobstructive azoospermia. All men were azoospermic and at least six years post-chemotherapy at the time of treatment. Eight of the 34 (24%) patients had also received extragonadal radiation. The mean interval from chemotherapy to TESE was 16.3 (range 6-34) years. Sperm were successfully retrieved in 46% (19/41) of attempts, with clinical pregnancy occurring in 37% (7/19) couples. For men with Hodgkin's disease, the sperm retrieval rate was 33% (4/12), whereas after treatment for germ cell tumor, the retrieval rate was 71% (5/7.) Diagnostic biopsy was not helpful in determining the prognosis for sperm retrieval. For men with Sertoli cell-only on diagnostic biopsy, the retrieval rate was 28% (5/18). No correlation was noted between the outcome of TESE-ICSI and the specific chemotherapeutic agents used.

Genetic testing for Y chromosome microdeletions is of prognostic significance for TESE procedures. For men with complete deletions of the AZFb region the chance of sperm retrieval during TESE is severely impaired. In our experience, 0 of 23 men with Y chromosome partial deletions involving all of AZFb had sperm retrieved with TESE, whereas the sperm retrieval rate in a contemporary series of men with non-obstructive azoospermia but no AZFb deletions was 67% (85/126). The presence of a complete deletion of AZFa appears consistently associated with Sertoli cell-only and a poor chance of sperm retrieval (Kamp et al., 2001). Of the men with complete deletions of the AZFa region, who had diagnostic biopsies or TESE at our institution, zero out of ten had sperm found. Therefore, for men with complete AZFa or AZFb deletions, we do not recommend proceeding with TESE (Hopps et al., 2003).

Pregnancy rates for ICSI using sperm from men with Y microdeletions appear to be very similar to those obtained for couples with similar sperm production. We recently reviewed a series of 27 IVF cycles involving men with AZFc microdeletions who were azoospermic (12 cycles) or severely oligospermic (15 cycles.) The clinical pregnancy rates per cycle were 57% (azoospermic � TESE cycles) and 33% (oligospermic). All children born were phenotypically normal, but we expect all boys to have deletions involving the AZFc region, with resulting impairments in spermatogenesis.

Should sperm retrieval be done in men who have sperm in the ejaculate?

In general, evidence from most ICSI studies has indicated that fertilization and pregnancy results are independent of sperm quality or number. However, more recent data has suggested that fertilization and pregnancy rates can be affected by sperm source or sperm quality (Palermo et al., 1999.) Several studies have shown that fertilization and pregnancy rates are lower when testicular sperm from men with non-obstructive azoospermia are used, when compared to results obtained with ejaculated sperm or epididymal sperm obtained by MESA from men with obstructive azoospermia (normal production). In addition, sperm from men with cryptozoospermia (so few sperm in the ejaculate that none are seen on an initial evaluation, but only after centrifugation and concentration of the sample) have lower fertilization and pregnancy rates than that obtained for men with higher numbers of spermatozoa in the ejaculate. Indeed, the results with cryptozoospermia are worse than those for men who have testicular sperm retrieved in non-obstructive azoospermia. Since most men with cryptozoospermia have sperm readily retrieved from the testis, we have offered testicular sperm retrieval to a select group of patients with cryptozoospermia who have had poor fertilization, embryo development and pregnancy results. We have also retrospectively analyzed results of ICSI for couples where rare sperm from the ejaculate were used for one cycle, but the man subsequently became azoospermic and had testicular sperm used for another cycle. Improved clinical pregnancy rates (64% vs. 41% per transfer) and embryo implantation rates (36% vs. 18%) are seen for these matched cycles when testicular sperm are used. An additional group of patients that may benefit from testicular sperm retrieval are the men with partial obstruction and high reactive oxygen species (ROS) activity in semen, for example men with very poor sperm motility after vasectomy reversal or those with partial ejaculatory duct obstruction. Again, anecdotal evidence in these patients suggests a better fertilization, implantation and pregnancy result when sperm are retrieved surgically.

We continue to use viable ejaculated sperm for any initial ICSI attempt at Weill Cornell rather than proceeding to surgical retrieval. However, for selected couples with cryptozoospermia, or evidence of partial obstruction, who have had poor fertilization and/or embryo development using ejaculated sperm in a prior cycle, sperm retrieval (usually by percutaneous needle biopsy) may be worthy of consideration.


Sperm retrieval is an effective adjunct for treatment of azoospermic men, including men with ejaculatory failure. At Weill Cornell, retrieval of sperm from men with ejaculatory failure or obstructive azoospermia is nearly certain. Using these sperm, pregnancy rates of 56% and 73% have been achieved for ejaculatory failure and obstruction, respectively, when advanced assisted reproduction, ICSI, is applied.

Sperm retrieval for use with ICSI, is now also possible for many men with non-obstructive azoospermia. Men with NOA may have unique genetic defects that should be evaluated prior to an attempt at conception. We have now attempted TESE for over 500 men during programmed IVF cycles at our institiution. At New York-Presbyterian Hospital-Weill Cornell Medical Center, sperm is retrieved from the testis with TESE in 57% of couples despite non-obstructive azoospermia, and 48% of couples achieve a clinical pregnancy using this sperm and ICSI. The chance of sperm retrieval in non-obstructive azoospermia is enhanced with the application of a microdissection technique. Since some couples will not have sperm retrieved with TESE, the potential use of frozen donor spermatozoa should be discussed with couples as a back-up, prior to simultaneous TESE-ICSI attempts. Couples in which men have obstructive azoospermia tend to have higher fertilization rates, and subsequent pregnancy rates than do couples in which men have non-obstructive azoospermia, in part because of the better genetic composition and quality of their sperm, a subject that will be discussed in later presentations.

Selected References

  1. Amer M, Ateyah A, Hany R, Zohdy W. Prospective comparative study between microsurgical and conventional testicular sperm extraction in non-obstructive azoospermia: follow-up by serial ultrasound examinations. Hum Reprod. 2000 Mar;15(3):653-6.
  2. Brandell RA, Mielnik A, Liotta D, Ye Z, Veeck LL, Palermo GD, Schlegel PN. AZFb deletions predict the absence of sperm with testicular sperm extraction: Preliminary report of a prognostic genetic test. Hum Reprod, 13:2812-2815, 1998.
  3. Choi JM, Chung P, Veeck L, Mielnik A, Palermo GD, Schlegel PN. AZF microdeletions of the Y chromsome and in vitro fertilization outcome. Fertil Steril 81:337-41, 2004.
  4. Chung PH, Palermo G, Schlegel PN, Veeck LL, Eid JF, Rosenwaks R. The use of intracytoplasmic sperm injection with electroejaculates from anejaculatory men. Hum Reprod 13:1854-1858, 1998.
  5. Decensi AU; Guarneri D; Marroni P; Di Cristina L; Paganuzzi M; Boccardo F. Evidence for testicular impairment after long-term treatment with a luteinizing hormone-releasing hormone agonist in elderly men. J Urol 142:1235-8, 1989.
  6. Devroey P, Liu J, Nagy Z, Goossens A, Tournaye H, Camus H, Van Steirteghem A, Silber S. Pregnancies after testicular sperm extraction and intracytoplasmic sperm injection in non-obstructive azoospermia. Hum Reprod 10:1457-1460, 1995.
  7. Fasouliotis SJ, Safran A, Porat-Katz A, Simon A, Laufer N, Lewin A. A high predictive value of the first testicular fine needle aspiration in patient s with non-obstructive azoospermia for sperm recovery at the subsequent attempt. Hum Reprod 17:139-42, 2002.
  8. Friedler S, Raziel A, Strassburger D, Soffer Y, Komarovsky D, Ron-El R. Testicular sperm retrieval by percutaneous fine needle sperm aspiration compared with testicular sperm extraction by open biopsy in men with non-obstructive azoospermia. Hum Reprod 12:1488-93,1997.
  9. Hopps CV, Mielnik A, Goldstein M, Palermo GD, Rosenwaks Z, Schlegel PN. Detection of sperm in men with Y chromosome microdeletions of the AZFa, AZFb and AZFc regions. Hum Reprod 18: 1660-1665, 2003 .
  10. Jaffe T, Oates RD. Genetic abnormalities and reproductive failure. Urol Clin No Am 21:389-408, 1994.
  11. Kahraman S, Ozgur S, Alatas C, Aksoy S, Tasdemir M, Nuhoglu A, et al. Fertility with testicular sperm extraction and intracytoplasmic sperm injection in non-obstructive azoospermic men. Hum Reprod 11:756-760, 1996.
  12. Lewin A, Reubinoff B, Porat-Katz A, Weiss D, Eisenberg V, Arbel R, Bar-el H, Safran A. Testicular fine needle aspiration: the alternative method for sperm retrieval in non-obstructive azoospermia. Hum Reprod 14:1785-90, 1999.
  13. Liotta D, Hariprashad J, Schlegel P, Zaninovic N, Veeck L. Do Severely oligozoospermic patients benefit from using testicular sperm rather than ejaculated sperm with ICSI? Abstract () presented at the ASRM 58th Annual Meeting, Seattle, Washington, October 12-17, 2002.
  14. Novero V, Goossens A, Tournaye H, Silber S, Van Steirteghem AC, Devroey P: Seminoma discovered in two males undergoing successful testicular sperm extraction for intracytoplasmic sperm injection. Fertil Steril 65:1051-4, 1996.
  15. O'Connell M, McClure N, Lewis SE. Mitochondrial DNA deletions and nuclear DNA fragmentation in testicular and epididymal human sperm. Hum Reprod. 2002 Jun;17(6):1565-70.
  16. Okada H, Dobashi M, Yamazaki T, Hara I, Fujisawa M, Arakawa S, Kamidono S.Conventional versus microdissection testicular sperm extraction for nonobstructive azoospermia. J Urol. 2002 Sep;168(3):1063-7.
  17. Okubo K, Ogura K, Ichioka K, Terada N, Matsuta Y, Yoshimura K, Arai Y, Honda T, Umeoka K, Nakahori T, Takahashi A, Umaoka Y. [Testicular sperm extraction for non-obstructive azoospermia: results with conventional and microsurgical techniques]. Hinyokika Kiyo. 2002 May;48(5):275-80
  18. Ostad M, Liotta D, Ye Z, Schlegel PN. Testicular sperm extraction (TESE) for non-obstructive azoospermia: Results of a multi-biopsy approach with optimized tissue dispersion. . Urology, 52:692-7, 1998.
  19. Palermo GD, Schlegel PN, Colombero LT, Zaninovic N, Moy F, Rosenwaks Z. Aggressive sperm immobilization in ICSI using epididymal and testicular spermatozoa improves fertilization and pregnancy rates. Hum Reprod 11:1023-1029, 1996.
  20. Palermo GD, Schlegel PN, Hariprashad JJ, Ergun B, Mielnik A, Zaninovic N, Veeck LL, Rosenwaks Z. Fertilization and pregnancy outcome with intracytoplasmic sperm injection for azoospermic men. Hum Reprod 14:741-748, 1999.
  21. Palermo GD, Schlegel PN, Sills ES, Menendez S, Veeck LL, Rosenwaks Z. Births following testicular sperm used with intracytoplasmic sperm injection in non-mosaic Klinefelter's syndrome. N Engl J Med 338: 588-560, 1998.
  22. Pavlovich CP, King P, Goldstein M, Schlegel PN. Evidence for a treatable endocrinopathy in infertile men. J Urology 165:837-841, 2001.
  23. Raman J, Schlegel PN. Aromatase inhibitors for male infertility. J Urol 167:624-629, 2002
  24. Ron-El R; Strassburger D; Friedler S; Komarovski D; Bern O; Soffer Y; Raziel A. Extended sperm preparation: an alternative to testicular sperm extraction in non-obstructive azoospermia. Hum Reprod 12:1222-6, 1997.
  25. Rosenlund B, Kvist U, Ploen L, Lundh-Rozell B, Sjoblom P and Hillensjo T. A comparison between open and percutaneous needle biopsies in men with azoospermia. Hum Reprod 13:1266-71,1998.
  26. Rucker GB, Mielnik A, King P, Goldstein M, Schlegel PN. Preoperative screening for genetic abnormalities in men with non-obstructive azoospermia prior to testicular sperm extraction. J Urol 160:2068-2071, 1998.
  27. Schlegel PN, Girardi SK: Clinical review: In vitro fertilization for male factor infertility. J Clin Endocrinol Metab, 82:709-716, 1997.
  28. Schlegel PN, Kaufmann J. The role of varicocelectomy in men with non-obstructive azoospermia. Fertil Steril 81:1585-8, 2004
  29. Schlegel PN, Palermo GD, Goldstein M, Menendez S, Zaninovic N, Veeck LL, Rosenwaks Z. Testicular sperm extraction with ICSI for non-obstructive azoospermia. Urology, 49:435-440, 1997.
  30. Schlegel PN. Su L-M. Physiologic consequences of testicular sperm extraction. Hum Reprod 12:1688-1692, 1997.
  31. Schlegel PN. Testicular sperm extraction: Microdissection improves sperm yield with minimal tissue excision. Hum Reprod, 14:131-135, 1999.
  32. Sheynkin Y, Schlegel PN. Sperm retrieval for assisted reproductive technologies. Contemporary Urology 9:21-36, 1997
  33. Silber SJ; Nagy Z; Devroey P; Tournaye H; Van Steirteghem AC: Distribution of spermatogenesis in the testicles of azoospermic men: the presence or absence of spermatids in the testes of men with germinal. Hum Reprod Nov;12(11):2422-8, 1997.
  34. Su LM, Palermo GD, Goldstein M, Veeck LL, Rosenwaks Z, Schlegel PN. Testicular sperm extraction with intracytoplasmic sperm injection: Preoperative histology can predict success of sperm retrieval. J Urol 161:112-116, 1999.