Pathogenesis of testicular carcinoma in situ and germ cell cancer: still more questions than answers

Testicular cancer is a disease of a great relevance for andrologists because young men at risk of this malignancy are frequently seen at andrology and fertility clinics. This special issue of International Journal of Andrology contains papers on testicular cancer: most of these papers were presented at the 7th Copenhagen Workshop on Carcinoma in Situ Testis and Germ Cell Cancer, held in October 2010. We have just marked a milestone of 30 years since the 1st Workshop in this series took place at the University Hospital (Rigshospitalet) in the Danish capital city. On the occasion of the 7th Workshop, we would like to take this opportunity to reflect on the legacy of previous meetings in this conference series and put into perspective the newest findings that are presented in this special publication. The 1st meeting was organised in 1980 in an attempt to answer some important questions concerning the biology of testicular germ cell tumours (TGCTs). The most important question was: does the carcinoma in situ (CIS) testis - first reported several years earlier (Skakkebæk 1972) - exist? Although not all participants at the conference in 1980 were convinced of its existence, a wealth of data later showed that the CIS testis is a forerunner of both seminomas (except spermatocytic seminoma) and non-seminomas. An important clinical implication of this discovery was that CIS can be detected by a testicular biopsy, which has become a common diagnostic tool (Berthelsen & Skakkebæk 1981). These data were summarized at the 2nd Copenhagen Workshop in what is now considered a classic paper (Skakkebæk et al., 1987). The new theory of TGCT was not yet final and shortly thereafter a correction was reported: infantile TGCTs that manifest as yolk sac tumours or benign teratomas have a different pathogenesis and their origin is still obscure (Manivel et al., 1988). Important issues regarding the biology and management of CIS were discussed at the 3rd Copenhagen Workshop, held in 1992. Novel markers for CIS, such as placental-like alkaline phosphatase (Jacobsen & Nørgaard-Pedersen 1984) were recently introduced in the clinical practice and a biologically important protein, KIT was detected in CIS (Rajpert-De Meyts & Skakkebæk 1994). Most scientists present at the meeting agreed that taking a contralateral biopsy and performing an immunohistochemical screen with CIS markers, should be offered as a service to all patients with a TGCT in order to prevent the development of bilateral cancers (Berthelsen et al., 1982). This new diagnostic method was well received especially in light of the promising results of treatment of CIS by radiotherapy, a novel approach that for the first time opened a possibility to prevent overt testicular cancer by eliminating the preinvasive stage (von der Maase et al., 1986; Giwercman et al., 1993). The last decades of the 20th century saw a rapid increase in the incidence of testicular cancer and other reproductive disorders in many countries around the world. This trend was met with a growing interest and many suggestions of possible explanations, including the so-called testicular dysgenesis syndrome (TDS) hypothesis (Skakkebaek et al., 2001). At the 4th and 5th Workshops the two main topics discussed were the possible aetiological role/s of environmental factors (Skakkebaek et al., 1998) and the developmental delay of germ cell differentiation as a common event in the pathogenesis of CIS-derived TGCTs (Rajpert-De Meyts et al., 1998, Rajpert-De Meyts 2006). An alternative theory of pathogenesis of TGCTs stipulating the oigin of CIS from spermatocytes was also presented and discussed (Chaganti & Houldsworth, 1998). Genomic aberrations, especially the implications of the consistent gain of chromosome 12p in TGCTs (Atkin & Baker 1982), was another important topic (Looijenga et al., 2003a; Skotheim & Lothe, 2003). Subsequent years witnessed an explosive growth in basic research on testicular cancer, mainly owing to the availability of new high-throughput methods of molecular analysis, an increase of interest in stem cell biology in general, and a concomitant increase of interest in the regulation of pluripotency in particular (Looijenga et al., 2003b; Sperger et al., 2003; Almstrup et al., 2004; Clark et al., 2004; Skotheim et al., 2005; Korkola et al., 2006). This shift in research focus was reflected in the programme for the 6th Workshop, which was mainly devoted to basic aspects of germ cell cancer, with emphasis on regulation of gene expression, including the emerging field of miRNAs (Looijenga et al., 2007; Rajpert-De Meyts, 2007). New pluripotency-related markers represented a good example of translational medicine as they helped re-instigate an older approach involving the non-invasive detection of CIS cells in semen (Hoei-Hansen et al., 2007). Looking back over the last 30 years there is little doubt that data from clinical medicine, epidemiology and environmental medicine, as well as molecular biology and genetics have improved our understanding of testicular cancer. During this time several new questions have arisen, for example: Should we strive to detect and treat CIS or would it be better to focus on treatment of overt TGCTs? Why has the incidence of TGCTs continued to increase in some countries but reached a plateau in others and why are there conspicuous national and geographical differences? Are environmental factors involved, and if so, what is the relative role of genetic factors? Does TDS exist and is CIS always derived from foetal gonocytes? What can we learn from genome-wide association studies? What is specific to the biology of germ cells in comparison to somatic cells in the testis that renders them so sensitive to neoplastic transformation? And, has progress been made in the search for experimental animal models and/or useful in vitro methods? Most of these questions were addressed at the 7th Workshop and are discussed in the ensuing papers published in this special issue. We highlight here our subjective choice of some contributions. The clinical part of the workshop focused on different phenotypes of TDS, epidemiological trends, methods of CIS detection, and the late consequences of TGCTs and treatments. It was confirmed that disorders of sex differentiation (DSD) constitute the most severe form of TDS. Moreover, with progress in molecular genetics, gene mutations are now being detected in less severe cases of TDS that were previously considered to be idiopathic. These less severe cases, which often exhibit disturbances in the developmental niche of germ cells, may serve as good models for mechanistic studies of the pathogenesis of TGCTs (Looijenga et al., 2011). However, in the majority of sporadic TGCTs and mild TDS forms an environmental aetiology is suspected, as demonstrated by rapidly changing incidence trends. In some countries the upward trend has abated while in other countries the incidence has suddenly surged (Jørgensen et al., 2011). The environmental factors responsible for these trends are unknown. Evidence concerning the role of selected endocrine-disrupting chemicals reviewed at the workshop remains weak and non-conclusive (Cook et al., 2011), consistent with the multi-factorial aetiology of TGCTs and TDS. All in attendance at the workshop were in agreement as to the need for early detection of TGCTs in order to avoid the potential side-effects of this disease. There is still a need for well-performed and analysed testicular biopsies which are the preferred current method for early TGCT detection (Dieckmann et al., 2011; Oosterhuis et al., 2011). The presentation of a novel method of immunocytochemical staining of CIS cells in semen aided by digitalized image analysis gave a realistic perspective of the implementation of this test for evaluation of all men at risk of TGCTs (Almstrup et al., 2011). This new method awaits further improvements of sensitivity and independent testing in other laboratories. A difficult clinical problem is that a large proportion of TGCTs are spreading rapidly and/or are resistant to treatment, forcing very aggressive high-dose chemotherapy, often in combination with autologous stem cell transplantation (Koychev et al., 2011). As presented by several speakers, there have been substantial achievements in the understanding of the mechanisms of tumour invasion and metastasis, both at the genomic and cellular levels (Díez-Torre et al., 2011; Koychev et al., 2011; Looijenga et al., 2011; Gilbert et al., 2011), as well as in a mouse model (Zechel et al., 2011). Such studies give hope for a better outcome prediction but there is still a long way to go before we can offer individualised therapy for each patient. Two large genome-wide association studies performed independently in the USA and UK offered novel biological insights into the area of genetic predisposition to TGCTs. The results were discussed at the workshop by the scientists who conducted these studies and a comprehensive review of the subject is included in this issue (Turnbull & Rahman, 2011). An attempt to use these data for clinical risk prediction has been proposed in another paper, which suggests that a combination of genetic risk variants with clinical factors, such as cryptorchidism or subfertility, may in the future help to predict the risk of disease, but more research is needed before such a prediction model can be clinically useful (Kratz et al., 2011). The fields of epigenetic aspects of inheritance and epigenetic regulation of gene expression have blossomed in recent years. As reviewed by the Rotterdam group (Looijenga et al., 2011), global DNA methylation differs according to the level of TGCT differentiation with hypomethylation in CIS cells (Wermann et al., 2010). Recent data on chromatin modifications in CIS cells were presented by the Copenhagen group: these studies demonstrated remarkably permissive patterns associated with high transcription and proliferation of these cells, despite enhanced heterochromatinization (Almstrup et al., 2010; Bartkova et al., 2011). Not surprisingly, the pattern of chromatin modifications in CIS cells is very similar to that of foetal germ cells (Almstrup et al., 2010). The close phenotypic resemblance between CIS and gonocytes has important implications. To understand the process of malignant transformation of gonocytes deep knowledge of the regulation of germ cell specification and pluripotency as well as useful experimental models, are required. Novel insight into mechanisms of stem cell differentiation was given by P.W. Andrews, whose group produced a series of new antibodies which may be useful to define the steps of differentiation of embryonal tumours (Wright et al., 2011). New excellent data on the regulation of expression of pluripotency-related factors and transcription factor TFAP2C (AP-2γ) in developing germ cells and TGCTs was presented by the Rotterdam group (Looijenga et al., 2011; Gillis et al., 2011) and the Bonn group (Schäfer et al., 2011). Evidence was also presented that suggests that developmentally regulated TGFβ/BMP signalling pathways may be involved in the pathogenesis of TGCTs (Fustino et al., 2011; Young et al., 2011). Much of the recent progress in the work on pluripotency regulation and tumour progression has been stimulated by the availability of the first truly seminoma-like cell line, TCam-2 (Mizuno et al., 1993), which served as an experimental tool for a number of original studies (Nettersheim et al., 2011; Young et al., 2011; Gillis et al., 2011). By contrast, experimental studies of foetal cells are still hampered by methodological difficulties, but some progress has been made owing to the refinement of human tissue xenografting models (Mitchell et al., 2010). Studies of gene expression in TGCTs have continued steadily in recent years and numerous sets of data of variable quality have been published. There is need to organise and catalogue all these data. The first attempt of this kind is presented in this issue as a comprehensive meta-analysis performed by the Oslo group (Alagaratnam et al., 2011), who succeeded in assigning unique transcriptomes to various histological types. Distinct transcriptomic profiles were found within each histological type, despite very similar genotypes. The TGCTs of young adult men that are preceded by CIS are the most common tumour type but not the only one occurring in the human testis. Rare TGCTs of early childhood and spermatocytic seminoma of older men are distinct entities. The pathogenesis of the infantile tumours remains largely unknown despite some inroads into the understanding of their phenotypic features (Palmer et al., 2010; Fustino et al., 2011). By contrast, recent studies of spermatocytic seminoma, summarised at the Workshop by A.O.M. Wilkie, revealed oncogenic gain-of-function mutations in this tumour (Goriely et al., 2009) and provided important novel insights into the biology of spermatogonial cells’ renewal (Hofmann & Waheeb, 2011). TGCTs in dogs, which preferentially occur in older animals, may serve as models for further studies of spermatocytic seminoma (Bush et al., 2011). In conclusion, our understanding of the pathogenesis of testicular cancer has markedly improved in recent years, and this publication is a testament to this progress. However, many questions remain, among them probably the most important one concerns the aetiology of TGCTs and the relative roles of genetic vs. environmental or lifestyle factors. We invite our readers to scrutinize this entire issue and hope that you will find inspiration to conduct studies that will help answer the outstanding questions.