Hum. Reprod. Advance Access originally published online on August 3, 2006
Human Reproduction 2007 22(1):2-16; doi:10.1093/humrep/del279
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Histological evaluation of the human testis—approaches to optimizing the clinical value of the assessment: Mini Review
1 Prince Henry’s Institute 2 Andrology Australia, Monash University, Monash Medical Centre, Clayton, Australia 3 University Department of Growth and Reproduction, Rigshospitalet, Copenhagen, Denmark and 4 Monash Institute of Medical Research, Monash University, Monash Medical Centre, Clayton, Australia
5 To whom correspondence should be addressed at: Prince Henry’s Institute, PO Box 5152, Clayton, Victoria, 3168, Australia. E-mail: rob.mclachlan{at}princehenrys.org
 |
Abstract |
|---|
Top
Abstract
IntroductionTesticular biopsy is a crucial assessment in reproductive practice with diagnostic and prognostic importance for assisted reproductive technologies (ARTs) and risk of testicular neoplasia. Endocrine and genetic tests cannot reliably distinguish obstructive azoospermia (OA) from non-obstructive azoospermia (NOA) or predict recovery of mature spermatids by testicular sperm extraction (TESE). Currently, divergent histological reporting systems and the use of imprecise terminology seriously degrade the value of the literature on TESE recovery rates and hamper evaluation of treatments and research on genotype–phenotype relationships. The rising incidence of testis cancer and carcinoma in situ (CIS), especially in infertile populations, requires that every effort be made for its early detection. We provide a systematic approach to the histological classification of spermatogenic disorders and detection of CIS in adult patients. We evaluate a large consecutive series of bilateral biopsies from infertile men and report (i) the frequency of bilateral or discordant patterns that supports the use of bilateral biopsy for comprehensive evaluation and (ii) a high prevalence of mixed patterns, particularly within the hypospermatogenesis classification, that helps account for reported success of TESE. We propose a new diagnosis code for testicular biopsies that addresses the needs of ART clinicians and allows data storage and retrieval of value in clinical practice and research.
Key words: male infertility/spermatogenesis/TESE/testis biopsy/testis cancer
|
Introduction |
|---|
The utility of clinical tests depends upon their appropriate application, technical performance and the provision of accurate and insightful reports. Specialized andrological tests such as semen analysis, testicular biopsy and reproductive hormone assays were initially developed and performed by clinicians but are now performed by laboratories for which they are often not a main focus. Continued re-evaluation of the service requirements of clinicians is necessary to avoid the adoption of procedures aimed at improving throughput and reducing costs, but that does not best serve clinical decision-making. Examples of such phenomena have been the expedient adoption of multiplatform testosterone assays that show significantly divergent normal ranges (Wang et al., 2004
; Sikaris et al., 2005) and the use of inappropriately high serum FSH reference intervals that may lead to misdiagnosis of obstructive azoospermia (OA) rather than non-obstructive azoospermia (NOA) (Schoor et al., 2002; Sikaris et al., 2005).
The evolution of assisted reproductive technologies (ARTs) has seen new approaches in the assessment of semen quality and sperm function, and an emphasis on identifying motile sperm in semen, or of elongated spermatids in testicular tissue, for ICSI. Only ART-affiliated laboratories focus on these matters. Furthermore, quality assurance programmes for semen analysis are a relatively recent development (Mortimer et al., 1986; Björndahl et al., 2002). As endocrine tests do not always distinguish with certainty normal (OA) from profoundly impaired spermatogenesis (NOA) nor predict the recovery of mature spermatids for ICSI via testicular sperm extraction (TESE), testicular biopsy remains a key investigation by defining the status of spermatogenesis.
The rising incidence of testis cancer and carcinoma in situ (CIS) (Richiardi et al., 2004) especially in at risk subgroups such as infertile men (Skakkebæk et al., 2001; Dieckmann and Pichlmeier, 2004) requires strategies for early detection, ideally at the pre-invasive stage. Testis biopsy is important in the evaluation of men at risk of CIS or testicular cancer (Dieckmann and Pichlmeier, 2004) such as those with idiopathic infertility (Møller and Skakkebæk, 1999), prior cryptorchidism (Giwercman et al., 1989), a history of testicular neoplasia (Dieckmann and Loy, 1996) or suggestive features on ultrasound, such as an identified lesion or microlithiasis (Holm et al., 2001, 2003).
Re-evaluation of testis biopsies reported externally often reveals that the original report was deficient and that clinical care may have been improved by the use of a systematized approach. Poor fixation is frequently noted and it is clear from the report that the pathologist does not frequently report on testicular biopsies. Divergent histological reporting systems and the use of imprecise terms may at least partly account for variable sperm recovery rates that cloud the literature on sperm recovery rates by TESE in NOA. It was refreshing to see such issues recognized in a recent study in which the histological diagnoses of 113 azoospermic men revealed significant reporting discrepancies and that changes in clinical management may have occurred in 27% of cases (Cooperberg et al., 2005), thereby highlighting the need for a standardized approach to reporting.
In this review, we focus on the assessment of spermatogenesis and detection of CIS, while the reader is directed to an excellent review of testicular neoplasia (Rørth et al., 2000). We propose an approach to the application, processing, evaluation and reporting of human testicular biopsy in an attempt to assist laboratory service providers to better provide the information required by clinicians. Major points will be expanded through discussion of current issues in fertility practice.
|
History and evolution of testis biopsy methods and intepretation |
|---|
The clinical role and approach to testis biopsy in male infertility goes back many decades (Charny, 1940
; Hotchkiss, 1944), indeed back to the early 20th century when needle aspiration was discussed by Huhner (1928). For these early physicians, and perhaps ideally to this day, testis biopsy was considered an essential component of the clinician’s examination, akin to fundoscopy in a diabetic. In the post-war years, many descriptive papers appeared that elegantly described the cells of the normal testis from immaturity to adulthood, and the common patterns seen in male infertility (Sniffen, 1950). Nelson (1950) applied key terms such as ‘germ cell aplasia’, ‘arrest’ and ‘generalized fibrosis’, the last one representing in fact hypospermatogenesis of varying degree. Linkages were made between treatment prospects and histology, such as inevitable sterility in Klinefelter’s syndrome or possible reconstructive surgery in ductal obstruction. Attempts at quantitative description (Roosen-Runge et al., 1957) of the testicular compartments and germ cell content were made and applied to understanding the processes leading to spermatogenic failure, such as that associated with chemotherapeutic treatments. Finally, the descriptions of the kinetics of the process of spermatogenesis provided a framework to consider not only histology but also the hormonal regulation of spermatogenesis (Clermont, 1963; Heller and Clermont, 1964).
A clinically insightful paper by Levin (1979) touched on key issues that still resonate today. He noted how the place of testis biopsy had changed with genetic (karyotyping) and endocrinological (gonadotrophin assays and treatment) discoveries that in some cases allowed diagnosis and management without recourse to biopsy, but he argued that biopsy still had a strong place in clinical practice. Today, fertility practice is impacted by (i) ICSI and the identification of mature spermatids in NOA (Devroey et al., 1995; Tournaye et al., 1997; Raman and Schlegel, 2003), including in 
70% of men with Klinefelter’s syndrome (Palermo et al., 1998; Shiff et al. 2005), (ii) the recognized linkage between testis developmental biology, cancer predisposition and infertility (Skakkebæk et al., 2001) and (iii) the research effort aimed at understanding phenotype:genotype relationships in male infertility (Foresta et al., 2001). In all settings, a clear and systematic approach to testicular biopsy is essential.
Johnsen (1970) proposed a scoring system for quantitatively describing spermatogenesis. By assessing many (100) tubules, it took note of the heterogeneity between tubules by grading them between 1 and 10 according to the most advanced germ cell in the profile. A key assumption was that the progressive degeneration of the tubule invariably features the loss of its cell contents in a fixed order beginning with the most mature (spermatozoa), proceeding to the loss of spermatogonia, then to Sertoli cells. A registration of the most mature cell type present was taken as an index of tubule quality. All tubules were classed from 10 (normal) to 5 (spermatocytes as the most advanced cell) and finally to 1 (no germ or Sertoli cells).
Unfortunately, the Johnsen score is fundamentally flawed. First, its basic proposition of a progressive loss of germ cells from the most mature to the basal stages is incorrect. Our quantitative data published in 1972 clearly demonstrated that in biopsies where there were marked decreases in the number of late spermatids, there was also loss in the more basal cell types such as round spermatids, primary spermatocytes and even spermatogonia (de Kretser et al., 1972). The second flaw is the use (or more correctly misuse) of the ‘mean tubule score’. To be fair, Johnsen asked that this profile of scores be considered in a tabular form and described a ‘mean score’ across all tubules. To illustrate the problem with the latter parameters, consider three settings: (i) complete germ cell arrest (GCA) at the primary spermatocyte stage in all tubules, (ii) combined spermatogenic failure and obstruction with 50% Sertoli cell-only (SCO) profiles and 50% normal tubules and (iii) moderately severe hypospermatogenesis in a patient with NOA in which an even scatter of most advanced scores from 1 to 10 is present. All provide a mean score of 5.0, yet from modern perspective, their outlook and management are starkly different, specifically (i) a poor outlook with TESE requiring at least consideration of donor sperm as an option, (ii) consideration of ductal obstruction and repair or TESE [perhaps even by fine needle aspiration (FNA)] and (iii) an excellent prognosis using TESE.
Despite its inherent flaws, the Johnsen score is still widely reported, but rarely according to the detail of its original description. In advocating its abandonment (including such modifications as proposed by de Kretser and Holstein, 1976), we recognize that there are no other systematic methods available and that as a consequence, confusing and inconsistent data will continue to appear as authors ‘go their own way’. The lack of methodological description and validation required by journals is in contrast to the rigorous validation required for endocrine assays. Perhaps more importantly, there is no means of appreciating the effects of such unstructured approaches in clinical practice that may compromise patient care.
|
Evaluation of the testicular biopsy and the importance of uniform definitions of the pathological states |
|---|
It is essential that the observer be provided with tissue fixed in a manner that the cell types during spermatogenesis can be easily recognized. The pathologist must be able to recognize the individual germ cell types with certainty by a thorough understanding of the morphological features of individual germ cells (Figure 1). All too frequently in clinical practice, the value of testicular biopsy is diminished as a result of poor fixation and reporting by pathologists not thoroughly familiar with the reproductive tract. Formalin, which is the most commonly used fixative in pathology departments, should not be used for the relatively soft testicular tissue because it causes shrinkage artefacts and makes it difficult to detect different types of germ cells and CIS cells. We recommend Bouin’s fluid or similar fixatives. The pathologist must systematically proceed in a step-wise manner to evaluate and describe the following key features:
|
- The uniformity of the histological features, and, if heterogeneity is evident, an estimate of the proportion of tissue with a given histological pattern should be given.
- The diameter of the seminiferous tubules, the presence of a lumen and the presence or absence of fibrosis.
- Evaluation of the germ cells present within the seminiferous epithelium, specifically seeking to identify spermatogonia, primary spermatocytes, round spermatids and elongated, condensed spermatids representing stages Sd1 and Sd2 of Clermont (1963, Figure 1). Evaluation and comment should be made as to whether any decrease in germ cells is mild, moderate or severe, and if spermatogenesis proceeds to the Sd2 stage, i.e. completed spermatid differentiation, when the term ‘hypospermatogenesis’ should be applied.
- If spermatogenesis ceases at a specific stage, the pathologist must identify at what stage of germ cell development the arrest occurs, and the term ‘arrest’ should not be used unless there is no progression beyond this stage in any tubule in the entire biopsy.
- The presence of Sertoli cells in the epithelium should be sought and their features noted. Small Sertoli cell nuclei with small nucleoli are described as immature. The term ‘Sertoli cell-only syndrome (SCOS)’ should only be used if all the tubules in the biopsy show a total absence of germ cells, thus when tubules containing only Sertoli cells are found in a biopsy in which other tubules contain advanced spermatids, the overall diagnosis is hypospermatogenesis.
- If both germ cells and Sertoli cells are absent from the tubules and there is marked fibrosis and the accumulation of hyaline amorphous basement membrane-like material, the term ‘seminiferous tubule hyalinization’ should be used.
- The spermatogonial compartment of the testis should be carefully examined to determine whether cells with the features of CIS cells are present and, if present, their distribution and whether there is any extratubular spread should be noted.
- The intertubular tissue should be examined to identify the presence of Leydig cells and their features. Accumulation of Leydig cells into large clumps (micronodules) should be noted. Additionally, the presence of any increase of macrophages or other inflammatory cells should be sought and described. The latter may vary from focal infiltration (e.g. around a tubule with CIS) to an extensive inflammatory process such as granulomatous orchitis (granuloma).
- Finally, other abnormalities seen in association with disturbed spermatogenesis should be noted including the presence of microlithiasis (intra- or intertubular concretions, usually concentric and calcified to some degree), clumps of dysgenetic tubules with distorted shapes of tubule membranes, Leydig cells trapped within hyalinized basement membranes or unusual blood vessels thickened by proteinaceous deposits.
The use of the above methodology then allows the classification of spermatogenesis (Table I). However, it is essential to recognize that these do not indicate a specific pathogenetic mechanism. If these definitions are accepted, then a significant degree of clarity will emerge and it will be possible for accurate comparison of the results in differing studies.
|
It has been suggested that Klinefelter’s syndrome presents a diagnostic histological pattern featuring hyalinized tubules together with two types of tubules, one containing Sertoli cells with the features of adult Sertoli cells and the other those with features of immature Sertoli cells with occasional sex chromatin visible in their nuclei as described by Skakkebæk (1969)
and Frøland and Skakkebæk (1971). Others are of the view that such tubules can also be seen in karyotypically normal men, especially if sex chromatin is not visible (de Kretser, personal communication). Examples of the different pure patterns of spermatogenic impairment (Figure 2), mixed and variant phenotypes (Figure 3) and other abnormalities seen in association with these patterns (Figure 4) are shown. The prevalence of these phenotypes in our experience of the 534 consecutive men referred for bilateral testicular biopsies for investigation of infertility (Table II) demonstrates the relative rarity of pure phenotypes and the high frequency of hypospermatogenesis and mixed patterns.
|
|
|
|
Upon this background, one can consider the indications for testis biopsy (Table III), its technical performance and the generation of a report of optimal clinical value. We have chosen to expand on three issues of direct relevance to modern clinical practice and also the impact on accurate phenotypic description on clinical research into spermatogenic and developmental disorders, including CIS.
|
Relationship between histology and sperm retrieval at TESE
Of the numerous reports of TESE in NOA, correlations have been reported between successful spermatid isolation and histology based on biopsies prior to, or at the time of, TESE (Friedler et al., 1997
; Mulhall et al., 1997; Tournaye et al., 1997; Jezek et al., 1998; Amer et al., 1999; Silber, 2000; Seo and Ko, 2001; Friedler et al., 2002; Tsujimura et al., 2002). However, obtaining a clear overview of this relationship is difficult with authors concluding that histology is either of limited or major prognostic significance (Schulze et al., 1999; Jezek et al., 1998). In our view, the lack of clarity may well arise from inaccurate identification of germ cells and confusing categorization of biopsies. A minority of reports cite referenced approaches to histological evaluation such as that of Levin (1979) (Friedler et al., 1997; Tournaye et al., 1997) or the modified Johnsen score described by de Kretser and Holstein (1976) (Jezek et al., 1998; Schulze et al., 1999; Tsujimura et al., 2002). More often, the description of the histological method is very brief, lacks precision and clarity or is even reduced to assessment by ‘formal histology’.
SCOS
For such a tightly definable phenotype, prevalence figures amongst men with severe spermatogenic disorders undergoing TESE are surprisingly variable ranging from 27 to 68% (Tournaye et al., 1997; Tsujimura et al., 2002). Some authors choose to combine pure and mixed phenotypes presumably as precise descriptions are considered too tedious and/or unnecessary, an approach that will inevitably result in higher sperm recovery rates. For example, Silber (2000) stated that ‘a biopsy showing isolated tubules with a few spermatids that were otherwise SCOS was, for simplicity, defined as SCOS’ and then reported a 57% spermatid recovery rate at TESE, a substantially higher rate than the 16–33% in studies in which a distinction was explicitly made (Tournaye et al., 1997; Jezek et al., 1998; Amer et al., 1999; Schulze et al., 1999; Seo and Ko, 2001; Tsujimura et al., 2002). The term ‘partial SCOS’ is often used when a variable proportion of tubules show some germ cell development, even to the elongated spermatid stage (Figure 3A). This is a relatively common pattern in idiopathic spermatogenic failure that may evolve over time into a pure SCOS phenotype. It is important to recognize this as a variant of hypospermatogenesis that carries a better outcome for sperm retrieval than that seen with true SCOS. In our group of 534 consecutive infertile men, who underwent bilateral biopsies, we found 15.7% to have bilateral pure SCOS and a further 6.4% to have unilateral SCOS with hypospermatogenesis (including late spermatids) on the other side (Table II). In a further 2.8% of men, some tubules were seen to be devoid of germ cells but were mixed with spermatid-containing tubules thus giving the patient classification of hypospermatogenesis.
In order to optimize its clinical value, the term SCOS should only be applied to a universal pattern wherein no germ cells are seen in any profile (Figure 2). This pattern is not indicative of a common pathological process being seen in idiopathic infertility, in association with Y microdeletion, or as a result of orchitis, chemo- or radiotherapy or may represent an embryological failure of germ cell migration to the gonadal ridge. TESE is successful in about a quarter of subjects with strictly defined SCOS reflecting focal spermatogenesis that is not apparent on biopsy and due presumably to heterogeneity and sampling error. Successful recovery rates of 18–23% have also been reported in SCOS following chemo- and radiotherapy (Chan et al., 2001; Meseguer et al., 2003). The relationship between possible different aetiologies of SCOS and the morphological features of the Sertoli cells in SCOS has been raised (Anniballo et al., 2000), but as yet there are no data to support methods for the identification of favourable prognostic features in true SCOS.
GCA
This term (or alternatively ‘spermatogenic arrest’), as discussed earlier, should only be used when there is complete interruption of spermatogenesis uniformly in all tubules (Figure 2). GCA at the primary spermatocyte stage is readily recognized and was observed in our series to be bilateral in its uniform form in 4.1% and in a further 8.4% in a mixed variant with some tubules showing progression of spermatogenesis only up to spermatogonia and some partially all completely hyalinized (Table II). Pure or mixed patterns with maturation only up to the stage of spermatocytes were present in only one testis (with hypospermatogenesis on the other side) in 10.1 % of men (Table II). In a further 10.1% of men, bilateral hypospermatogenesis was present but some tubules showed a failure of germ cell progression beyond this spermatocyte stage. Thus, we find GCA to be less common than other studies, even when the distinction appears to have been sought between partial and pure variants (27–44%) (Friedler et al., 1997, 2002; Tournaye et al., 1997), but similar to the 6–16% reported by some (Amer et al., 1999; Schulze et al., 1999; Seo and Ko, 2001; Tsujimura et al., 2002). The term ‘partial GCA’ is used by some to mean either (i) a widespread and marked decline in germ cell progression across meiosis but still in association with a few elongated spermatids (Silber et al., 1996, 1997) or (ii) the presence of tubules showing either GCA or hypospermatogenesis (Figure 3B). We regard these as variants of hypospermatogenesis that consequently carry a markedly better outlook with TESE.
Spermatid recovery rates by TESE of 100% have been reported when the term GCA is loosely used (Silber et al., 1996), but more often recovery rates are between 14 and 33% (Friedler et al., 1997, 2002; Amer et al., 1999; Osmanagaoglu et al., 2003). More systematic evaluation of biopsies are needed to clarify the literature about the outlook for pure GCA at the primary spermatocyte or spermatogonial stages, but it appears to carry a worse outlook for successful TESE than SCOS, such that options such as donor insemination ought be discussed in the likely event of failed sperm isolation.
Hypospermatogenesis
The finding of a mature spermatid in any tubule profile points to the completion of spermatogenesis and defines the condition of hypospermatogenesis. The patterns can be uniform across tubules (Figure 2), but more often there is variability between tubules including some with SCO or apparent arrested germ cell development (Figure 3). The latter patterns are likely to be the true phenotype in many cases of successful TESE, described under the terms ‘partial’ SCOS or GCA. Such mixed patterns are common, but from an ART perspective, the most advanced tubule represented in the biopsy is well correlated with the success of TESE (Schulze et al., 1999) making it exceptionally important that the pathologist assiduously examines every tubule profile. In our group of 534 consecutive infertile men undergoing bilateral biopsies, hypospermatogenesis was present bilaterally (alone or mixed with tubules showing more severely impaired germ cell development) in 32%, in one testis in 17.6% and in a further 10.7% with discordant patterns resulting in hypospermatogenesis as the classification in about 60% of all cases. Occasionally, profiles of apoptotic cells can be confused with the heads of Sd2 spermatids (Figure 4D). Careful evaluation is necessary in these cases to determine if a sperm tail is attached to these pyknotic structures.
It is perplexing that sperm do not appear in the ejaculate despite the presence of elongated spermatids on diagnostic biopsy or at TESE, even using FNA that provides very little tissue (Craft et al., 1997). One unconvincing explanation is that there is a quantitative threshold of ‘spermatid number per tubule profile’ below which sperm are too scarce to be noted in semen. Thus, it has been proposed that four to six spermatids per tubule profile is associated with NOA in contrast to 17–35 in normal men (Silber et al., 1997), yet such a 4-fold difference could not account for 10 million-fold difference in sperm output. More likely is a defect in the last phases of spermiogenesis, including the process of sperm release (spermiation), in many men with NOA and hypospermatogenesis. The physiology of normal spermiation is poorly understood (Lee and Cheng, 2004) and is almost unstudied in the context of infertility.
Finally, the term ‘azoospermia’ asserts that sperm are absent from the ejaculate, but this cannot be objectively supported as ‘proving an absence’ is difficult. A recent evaluation of semen analysis procedures indicated that the lower limit of detection limit using the Neubauer chamber was around 10,000/ml and that centrifugation methods significantly underestimated sperm density (Cooper et al., 2006). Thus, like other laboratory parameters, the term ‘undetectable’ is more applicable. Intermittent azoospermia is a frequent finding in severe infertility, and such men have ‘intermittent and exceedingly severe oligoozospermia’ rather than NOA. This concept further reduces the mystery of successful TESE in NOA and aligns with the frequent detection of elongated spermatids in many severely infertile men, including those with Klinefelter’s syndrome (Schiff et al., 2005) (Figure 3F).
Differentiating OA versus NOA
The clinical features of OA include normal testicular volumes, serum FSH and inhibin B, with or without risk factors of obstruction at different anatomical levels. However, equivocal endocrine and/or clinical findings may dictate the need for biopsy. Although the distinction between OA and NOA is usually obvious, in some instances decreased sperm production, perhaps resulting from the obstructive lesion, may result in a pattern of hypospermatogenesis making the distinction between the two difficult. Chronic OA can result in reduced spermatogenesis, e.g. in vasectomy (Raleigh et al., 2004). Thus, a ‘semiquantitative’ appreciation comes with experience that there is a need to consider obstructive lesions for surgical correction.
Even when the diagnosis of obstruction is almost certain, it is sometimes necessary to confirm that spermatogenesis is occurring normally before undertaking expensive and invasive therapies such as ICSI or reconstructive surgery. In our experience, and that of others (Schulze et al., 1999), vasectomized men presenting for ICSI almost always have spermatids readily obtainable by needle aspiration and thus are not routinely checked pre-ICSI. After discussion of the low risk of failed TESE, this approach may also be reasonable in men with bilateral absence of the vas or cystic fibrosis, the latter being an anaesthetic risk of additional surgeries.
Determining the presence of testicular dysgenesis and CIS
The identification and management of co-existent illness in men presenting to ART programmes is essential as androgen deficiency (Andersson et al., 2004), testicular neoplasia and psychosexual/erectile problems are more frequent. A pathogenetic link has been hypothesized between testicular neoplasia and infertility related to impaired testicular development (Skakkebæk et al., 2001; Asklund et al., 2004). Thus, a history of cryptorchidism must be sought as the best-known risk factor of testicular cancer (Møller et al., 1996) and also as it is associated with impaired spermatogenesis even when timely orchidopexy has been performed. An association between impaired spermatogenesis and hypospadias is less well documented (reviewed in Mieusset and Soulie, 2005).
Even in the absence of such a history, subfertility appears to be a risk factor of testicular cancer (Møller and Skakkebæk, 1999; Raman et al., 2005) although the relative risk is low because of its heterogeneous aetiology (reviewed in Dieckmann and Pichlmeier, 2004). Fertility problems are common in men with testicular cancer. Sperm densities are often much poorer than might be expected if one functioning testis were present (Berthelsen and Skakkebæk, 1983). Men who subsequently develop testis cancer have had significantly fewer children than controls (Møller and Skakkebæk, 1999) and more often have abnormal semen quality (Jacobsen et al., 2000).
In our series of 534 consecutive men who underwent bilateral testicular biopsy because of fertility problems, CIS was detected in 13 men (2.4%, Table II), confirming that infertile men are at greater risk than the general male population in Denmark (the current life-time risk is 1.2%). The true prevalence of CIS/overt germ cell neoplasia among the infertile men is almost certainly greater as (i) men presenting with subfertility are not aware of this condition usually until their early thirties, while testicular cancer develops frequently in a younger age and (ii) testicular biopsies are performed after eliminating the possibility of a testicular mass by palpation and scrotal ultrasonographic examination.
Histological studies provide additional evidence that dysgenetic features, such as undifferentiated tubules or hyaline bodies, are relatively common among men referred to andrology and fertility clinics (Skakkebæk et al., 2003). Such dysgenetic features are frequently present in seemingly normal contralateral testis in patients with unilateral tumours (Hoei-Hansen et al., 2003). CIS cells, primordial germ cells and gonocytes are morphologically similar and share a common pattern of expression of various antigens (Jørgensen et al., 1995; Rajpert-De Meyts et al., 2003; Honecker et al., 2004). Studies of the regulation of the cell cycle in normal and neoplastic germ cells provided additional evidence that CIS cells have predominantly features of mitotically dividing immature germ cells (reviewed in Bartkova et al., 2003). More recently, genomewide expression studies confirmed the close phenotypic similarity between CIS cells and primordial germ cells/gonocytes, as both cell types express a number of genes associated with pluripotency typical for embryonic stem cells (Almstrup et al., 2004).
These data are highly pertinent to modern fertility practice as testis biopsy material must be assessed not only in regard to spermatogenesis but also with a mind to the possibility of co-existent CIS, or cancer. The histological evaluation requires familiarity with the morphological appearance of CIS cells and the use of immunohistochemical (IHC) labels that detect expression of primordial germ cell proteins (Figure 5). Given the prevalence of CIS in infertile men (1–5%) (Møller and Skakkebæk, 1999; de Gouveia Brazao et al., 2004; this series), it is striking that the literature focusing on TESE rarely if ever mentions this issue. It is unclear whether any assessments are being performed routinely, whether cases are being missed because of inadequate sampling, preparation and a failure to use IHC methods or whether pathologists are even alert to the association between CIS and infertility.
|
The mandatory collection of biopsy material for histology during evaluation or at the time of TESE has been recommended and CIS and incidental cancers have both been reported at TESE (Novero et al., 1996
; Schulze et al., 1999). We have made such routine assessments for many years and have found several cases of CIS. Despite the good prognosis of disseminated testicular cancer, the early detection of testis cancer at a pre-invasive stage is highly desirable as it reduces the extent of treatment, allows for sperm cryopreservation and enhances the prospect for continued function of the contralateral testis (reviewed in Hoei-Hansen et al., 2005).
Implications for clinical research
Clinical studies examining the effects of treatments or phenotype–genotype studies, such as those relating to chromosome Yq deletions, are handicapped unless an agreed system of histological evaluation is used. In this regard, uniform histological patterns such as GCA are particularly useful, as they would seem to reflect a uniform genetic or toxicant exposure. GCA at the primary spermatocyte stage, when found, may indicate the presence of a genetic lesion involved in the successful completion of meiosis. Thus, careful assessment of testicular biopsies may provide useful clues as to the genes that may require evaluation. Given the heterogeneity between tubules within men, and between men exposed to similar testicular insults, future studies need to consider the gene/protein expression studies within tubules using techniques such as microdissection of individual tubules.
|
Considerations in the number and type of biopsy |
|---|
Testicular biopsy can be performed under local or general anaesthesia and compromise either a trans-cutaneous needle (aspiration or core) or open biopsies from one or more sites (all either unilateral or bilateral). The clinical issue must be matched to the procedure, and it is recognized that some matters are both contentious and evolving. There are settings in which the primary clinical issue is readily addressed using FNA (e.g. differentiating obstructive from NOA or retrieving spermatids post-vasectomy), whereas open testicular biopsy is preferred in assessing a subject at high risk of CIS or when contralateral testis cancer is present. Other settings need to be considered on a case-by-case basis including pragmatic issues such as the cost and availability of services.
FNA
FNA is a simple, low-cost and low-risk procedure that, despite some induction of artefacts, provides a histological result that is quite adequate for differentiating between OA and NOA and for assessing spermatogenic defects, as it shows good agreement with open biopsy data (Mallidis and Baker, 1994) (Figure 4D). At least 20 tubule profiles are identified, which is adequate in most clinical settings (Mallidis and Baker, 1994; Craft et al., 1997; Tuuri et al., 1999). For sperm retrieval, FNA readily yields sperm for ICSI in OA with minimal cost and morbidity (Tuuri et al., 1999) and may also provide sperm in NOA, particularly in the setting of hypospermatogenesis (Mercan et al., 2000). Recovery rates are significantly lower than with multiple open biopsies (Friedler et al., 1997), whereas microdissection TESE (Schlegel, 1999; Tsujimura et al., 2002) appears to offer the highest retrieval rates. We have used a combined approach by first undertaking an FNA in those with favourable diagnostic histology (e.g. hypospermatogenesis) if this is available, but to schedule an open TESE procedure to follow immediately should this fail (‘FNA with open back up’).
Open testicular biopsy
Open testicular biopsy of at least 3 x 3 x 3 mm provides many more tubule profiles and regional representation and thus a higher probability of revealing a mixed phenotype that may then alter the clinical approach. We have observed markedly different histological patterns between sides in 28% of our population and a high incidence of mixed patterns (Table II) lending support to the use of bilateral biopsies in providing a more thorough evaluation that may assist clinical decision-making. We recommend an open surgical biopsy of the contralateral testis at the time of surgery for the primary unilateral testicular cancer, in order to exclude the presence of CIS. Added benefits include the excellent assessment of the spermatogenic/fertility potential of the other testicle and the patient’s peace of mind concerning the possibility of bilateral cancer.
CIS is often generalized and identifiable even within the small piece of tissue provided by FNA, but existing detection strategies are based on open biopsies. We advocate performing open biopsy in all cases with some suspicion of CIS (e.g. testicular atrophy, microlithiasis) with a single biopsy per testis although others have argued for at least two, despite the increasing risk of testicular damage (Loy et al., 1990; Kliesch et al., 2003; Dieckmann and Pichlmeier, 2004).
Cryopreservation of spermatids from TESE for subsequent ICSI provides a means of both avoiding futile ART cycles and of reducing the number of TESE procedures (Oates et al., 1997; Ben Yosef et al., 1999; Verheyen et al., 2004). However, it is predicated on the successful thawing of viable sperm and carries the risk of needing repeat (potentially perhaps unsuccessful) TESE procedures with their concomitant risks (Verheyen et al., 2004). This practice is clearly evolving, but in its ongoing evaluation, consideration should be given to the fact that it would provide a wider systematic and detailed histological evaluation of spermatogenesis and CIS.
|
Approach to laboratory processing and reporting of testicular biopsy |
|---|
A systematic approach to the provision of a high-quality slide for histological analysis (Table IV) is based on the combined experience of our two centres and the previous suggestions and observations, e.g. concerning surgical complications of urologists (Bruun et al., 1987
; Dieckmann et al., 2003). Such guidelines would be readily followed within any pathology department, but special attention (e.g. the avoidance of formalin fixation) does require a ‘non-standard’ approach be adopted.
|
Methods that help to identify CIS cells
It is not always easy, even for an experienced pathologist, to identify CIS cells, especially in formalin-fixed biopsies; therefore, it is advisable to include a serial section stained by immunohistochemistry for a CIS marker, e.g. placental-like alkaline phosphatase (PLAP, the most commonly used, an example is shown in Figure 5), M2A (D2–40), OCT-3/4, KIT or AP-2
. Other identification methods include periodic acid Schiff (PAS) or Laczko-Lévai (in semithin sections) staining for glycogen stored in CIS cells (Rørth et al., 2000).
Grading system
We suggest use of a descriptive system, conveniently coded for laboratory reporting and data recovery purposes, that is specifically designed to be relevant to clinical planning and research (Table V). This approach is predicated on first describing the type of testis (adult, immature or neoplasia), then the main (or only) spermatogenic pattern and then the second most prevalent pattern followed by other features. Significant additional features outside this classification system, such as the presence of multinucleated or degenerating germ cells, or vacuolation of Sertoli cells, should be mentioned in the descriptive part of the report. Data from each testis are recorded separately. In our department, several digital images are stored for ready retrieval and interesting features noted to facilitate identification and retrieval of samples of interest.
|
|
Generation of reports, data retrieval and conclusion |
|---|
Most male infertility now is seen in association with ART programmes and is channelled through both gynaecologists and urologists. It is essential that clinician specialists engaged in larger ART programmes interact with their pathology departments and develop systems tailored to their specific needs. The reports generated from our system can be partially automated but also allow for the description of additional features and perspectives to be added manually. For example, a report coded ‘1222’ would be reported as ‘Moderate severe hypospermatogenesis with a general reduction in advanced elongated spermatids and interstitial fibrosis’ but with the added comment: ‘The recovery of elongated spermatids for ICSI is likely using FNA with open biopsy back up’. Such tailored reports require that the clinician and pathologist are intimately familiar with the ART practice. For smaller units and other settings, the reporting testis biopsies by a central reference laboratory may be appropriate.
This classification approach is an invaluable audit and research tool. As summarized in Table II, data retrieval is readily performed and could be tailored to relate to any number of clinico-pathological correlative studies such as ART outcomes, genotype–phenotype studies or following of clinical or endocrine end-points.
|
Acknowledgements |
|---|
This work was supported by grants from the Danish Medical Research Council, the Danish Cancer Society and The Svend Andersen Foundation and by an Australian National Health and Medical Research Council Program Grant (#241000) and Fellowship (RMcL, #169020).
|
References |
|---|
Almstrup K, Hoei-Hansen CE, Wirkner U, Blake J, Schwager C, Ansorge W, Nielsen JE, Skakkebæk NE, Rajpert-De Meyts E, Leffers H. (2004) Embryonic stem cell-like features of testicular carcinoma in situ revealed by genome-wide gene expression profiling. Cancer Res 64:4736–4743.
Amer M, Haggar SE, Moustafa T, Abd El-Naser T, Zohdy W. (1999) Testicular sperm extraction: impact of testicular histology on outcome, number of biopsies to be performed and optimal time for repetition. Hum Reprod 14:3030–3034.
Andersson A-M, Jørgensen N, Larsen LF, Rajpert-De Meyts E, Skakkebæk NE. (2004) Impaired Leydig cell function in infertile men: a study of 357 idiopathic infertile men and 318 proven fertile controls. J Clin Endocrinol Metab 89:3161–3167.
Anniballo R, Ubaldi F, Cobellis L, Sorrentino M, Rienzi L, Greco E, Tesarik J. (2000) Criteria predicting the absence of spermatozoa in the Sertoli cell-only syndrome can be used to improve success rates of sperm retrieval. Hum Reprod 15:2269–2277.
Asklund C, Jørgensen N, Jensen TK, Skakkebæk NE. (2004) Biology and epidemiology of testicular dysgenesis syndrome. BJU Int 93:6–11.[ISI][Medline]
Bartkova J, Rajpert-De Meyts E, Skakkebæk NE, Lukas J, Bartek J. (2003) Deregulation of the G1/S-phase control in human testicular germ cell tumours. APMIS 111:252–266.[CrossRef][ISI][Medline]
Ben Yosef D, Yogev L, Hauser R, Yavetz H, Azem F, Yovel I, Lessing JB, Amit A. (1999) Testicular sperm retrieval and cryopreservation prior to initiating ovarian stimulation as the first line approach in patients with non-obstructive azoospermia. Hum Reprod 14:1794–1801.
Berthelsen JG and Skakkebæk NE. (1983) Testicular cancer: abnormal structure and function of the contralateral testis. Int J Androl 6:209–211.[ISI][Medline]
Björndahl L, Barratt CL, Fraser LR, Kvist U, Mortimer D. (2002) ESHRE basic semen analysis courses 1995–1999: immediate beneficial effects of standardized training. Hum Reprod 17:1299–1305.
Bruun E, Frimodt-Møller C, Giwercman A, Lenz S, Skakkebæk NE. (1987) Testicular biopsy as an outpatient procedure in screening for carcinoma in situ: complications and the patient’s acceptance. Int J Androl 10:199–202.[ISI][Medline]
Chan PT, Palermo GD, Veeck LL, Rosenwaks Z, Schlegel PN. (2001) Testicular sperm extraction combined with intracytoplasmic sperm injection in the treatment of men with persistent azoospermia postchemotherapy. Cancer 92:1632–1637.[CrossRef][ISI][Medline]
Charny CW. (1940) Testicular biopsy. Its value in male sterility. J Am Med Assoc 115:1429–1432.
Clermont Y. (1963) The cycle of the seminiferous epithelium in man. Am J Anat 112:35–45.[CrossRef][ISI][Medline]
Cooper TG, Hellenkemper B, Jonckheere J, Callewaert N, Grootenhuis AJ, Kersemaekers WM, Leung A, Wang C. (2006) Azoospermia: virtual reality or possible to quantify? J Androl 27:483–490.
Cooperberg MR, Chi T, Jad A, Cha I, Turek PJ. (2005) Variability in testis biopsy interpretation: implications for male infertility care in the era of intracytoplasmic sperm injection. Fertil Steril 84:672–677.[CrossRef][ISI][Medline]
Craft I, Tsirigotis M, Courtauld E, Farrer-Brown G. (1997) Testicular needle aspiration as an alternative to biopsy for the assessment of spermatogenesis. Hum Reprod 12:1483–1487.
de Kretser DM and Holstein AF. (1976) Testicular biopsy and abnormal germ cells. In Hafez E.S.E. (Ed.). The Human Semen and Fertility Regulation in Men(Mosby Co., St. Louis, Missouri) pp. 332–343.
de Kretser DM, Burger HG, Fortune D, Hudson B, Long AR, Paulsen CA, Taft HP. (1972) Hormonal, histological and chromosomal studies in adult males with testicular disorders. J Clin Endocrinol Metab 35:392–401.[ISI][Medline]
Devroey P, Liu J, Nagy Z, Goossens A, Tournaye H, Camus M, van Steirteghem A, Silber S. (1995) Pregnancies after testicular sperm extraction and intracytoplasmic sperm injection in non-obstructive azoospermia. Hum Reprod 10:1457–1460.[ISI][Medline]
Dieckmann KP and Loy V. (1996) Prevalence of contralateral testicular intraepithelial neoplasia in patients with testicular germ cell neoplasms. J Clin Oncol 14:3121–3125.[Abstract]
Dieckmann KP and Pichlmeier U. (2004) Clinical epidemiology of testicular germ cell tumors. World J Urol 22:2–14.[CrossRef][ISI][Medline]
Dieckmann KP, Classen J, Loy V. (2003) Diagnosis and management of testicular intraepithelial neoplasia (carcinoma in situ)—surgical aspects. APMIS 111:64–68.[CrossRef][ISI][Medline]
Foresta C, Moro E, Ferlin A. (2001) Prognostic value of Y deletion analysis. The role of current methods. Hum Reprod 16:1543–1547.
Friedler S, Raziel A, Strassburger D, Soffer Y, Komarovsky D, Ron-el R. (1997) 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 12:1488–1493.