Abstract
Testicular cancer is the most common cancer in young adult men, and its incidence is increasing globally, with testicular germ cell tumors (TGCT) being the most frequent subtype. These tumors show remarkable sensitivity to cisplatin, but there is a small subset of patients that develop resistance to the therapy, experiencing low quality of life and having no treatment options. Nowadays, there are a number of models that are useful for studying TGCT tumorigenesis and identifying novel targeted therapies for these patients. Thus, in this review, we aim to provide a comprehensive overview of the in vitro and in vivo TGCT models that are currently available, and also to summarize the main findings that have been made in the last years regarding preclinical and clinical studies of novel targeted therapies in TGCT patients.
Plain language summary
Testicular cancer is the most common type of cancer in young adult men. Most patients have a good prognosis due to the effectiveness of chemotherapy, which usually eliminates these cancer cells efficiently. However, there is a small percentage of patients that develop resistance to chemotherapy, and there are no effective alternative treatment options for such cases. Thus, there is a need for new therapies for those patients. In this review, the authors describe the laboratory models currently available to study this disease and how they can be used to test new drugs. Also, novel therapies that are being tested in testicular cancer patients are also discussed, to provide a comprehensive summary of both laboratory and clinical research being performed worldwide.
Keywords
Introduction
Testicular cancer is the most common malignancy in young men (15–40 years old) worldwide. 1 Testicular germ cell tumors (TGCT) are the most frequent type and make up about 98% of all testicular malignancies. 1 Their etiology is still widely unknown. Major TGCT risk factors include genetics, with genome-wide association studies (GWAS) identifying variants in over 70 genes associated with TGCT risk, developmental abnormalities such as cryptorchidism, and environmental exposures to particular chemicals like pesticides.2,3 Certain occupations associated with environmental exposure such as firefighters and military servicemen also have elevated tumor risk.4,5
TGCTs are mainly classified into seminomas (SE) and nonseminomatous tumors (NST), and SE are slightly more common than NST. The latter are divided into embryonal carcinoma (EC), choriocarcinoma (CH), yolk sac tumor (YST), and teratoma (TE) histologies depending on differentiation, and can be pure or (more frequently) mixed forms (i.e., combinations of two or more histologic subtypes). 1
Testicular cancer is a model of curable malignancy, with about 70%–75% of all cases diagnosed at clinical stage I, 6 of which 70%–85% are curable with orchiectomy alone. 7 Furthermore, TGCTs represent a model of success of platinum-based chemotherapy treatments, namely cisplatin. This compound was implemented in the 1970s and rapidly became standard treatment for these patients due to TGCTs remarkable sensitivity. Cisplatin use has increased the 5-year survival rate to 95%, even in the event of metastatic disease. 8 However, and despite the obvious advantages of such treatment, this success comes with the price of short- and long-term toxicity in these young individuals who survive their cancer. 9 Furthermore, around 10%–20% of TGCT patients experience relapse and about half of those ultimately develop resistance to cisplatin treatment, even after attempts at high-dose salvage therapy and stem cell transplant, with very high morbidity and mortality. 10 In this setting, there is a lack of therapeutic alternatives with clinical efficacy. Overall, there is a need for novel treatment strategies that may be combined with cisplatin, allowing to lower its effective doses or re-sensitize tumors to the compound, or that can replace it and target alternative mechanisms at play when true platin-resistance occurs.
Drug discovery and development are very challenging processes that take several years, from compound discovery and preclinical models (in vitro and in vivo) to clinical trials and drug approval for cancer treatment. 11 Generally, drug development can be divided into several stages: prediscovery, drug discovery, preclinical studies, clinical trials, and, finally, reviewing and approval. All these stages need to be sequentially accomplished and verified until approval, and on average, 12 years are required from the beginning of preclinical studies to drug approval. 12 Furthermore, only about 10%–20% of compounds tested in clinical trials receive market approval from regulatory agencies, 13 evidence that this is a thorough and slow process.
Thus, it is very important to perform highly accurate and reproducible preclinical studies with novel compounds that have the potential to move into clinical trials. For this to happen, appropriate in vitro and in vivo TGCT models are warranted, to mimic as closely as possible the human tumor setting and its microenvironment. Preclinical cellular and animal models serve as platforms to screen and analyze new agents’ anti-tumor potential, toxicity, pharmacokinetics, and efficacy. In general, in vitro models are the first models used to demonstrate drug potential in tumor cells, and animal models are ultimately employed in a more advanced stage to determine efficacy in a physiological setting as well as to assess compound toxicity/safety and metabolism, so that drugs can be safely applied in human clinical trials. Ultimately, innovative therapies are needed to bestow clinicians with the means to properly address clinical challenges in TGCTs, namely the major issues of young patients’ overtreatment and cisplatin resistance.
In this review, we provide a detailed overview of available preclinical models that may aid researchers to facilitate drug development and testing specifically for TGCT patients. Furthermore, novel targeted therapies that are currently in clinical trials and/or entering the routine clinical practice for TGCT patient treatment are also covered.
Methods
To perform this review, PubMed was searched for articles published between 1954 and 2025 with the terms “testicular cancer,” “testicular tumor,” or “testicular germ cell tumors” (36,875 articles in total for the initial screening). In order to capture the present scenario, clinical trials within the past 35 years were prioritized. This way, we encompassed all the relevant trials that used targeted therapies for the treatment of these tumors. Additionally, the search was narrowed to list the known in vitro and in vivo experimental models of testicular tumors that are used in preclinical translational research. A traditional narrative review approach (although more limited that a formal systematic review) was used in the writing, in order to comprehensively summarize the most relevant findings in the field, while also commenting on strengths and weaknesses of each model.
Preclinical models
In vitro models
Since the first establishment of a cancer cell line in 1953 (HeLa cells 14 ), in vitro models have been extensively used in cancer research. They provide almost infinite biological material for research purposes, and although not being totally representative of their tumors of origin, 15 are very useful models for initial preclinical studies. TGCT cell lines are widely used in preclinical experiments to study targeted therapies’ effects. There are currently both primary and metastatic TGCT cell lines available to properly build experiments for different settings. 16 Furthermore, many of these cell lines have cisplatin-resistant counterparts developed by various groups, 17 as cisplatin resistance remains one of the major unsolved clinical issues for TGCT patients. These cisplatin-resistant clones are known to have transcriptomic differences when compared with their parental cell lines, in genes that may be related with the resistance mechanisms.18,19 TP53 is one such gene, which has been implicated in cisplatin resistance, along with its regulator MDM2. 20 Primary TGCTs are frequently wild-type for TP53, such as the case of the 2102-Ep cell line, whereas mediastinal GCT, like NCCIT which is hemizygously mutated TP53, are frequently mutated for TP53. 20 Thus, these cell lines are vital to study the biological processes of cisplatin resistance (which is multifactorial in nature) and to uncover novel targets that may be therapeutically actionable to bypass it. However, the number of available cell lines is limited 21 (about 25, represented in Table 1), and they fail to recreate tumor heterogeneity, as a significant percentage of TGCTs are mixed tumors, which is a feature that is difficult to mimic in culture systems. Furthermore, although SE makes up for almost half of TGCTs, most TGCT cell lines are representative of EC phenotype. The only cell line with confirmed SE characteristics is TCam-2, which shows isochromosome 12p along with mRNA, microRNA, and specific protein markers expression consistent with clinical SE, although presenting with a BRAF gene mutation, which likely explains the reason why these cells could be propagated in vitro. 22 Although a second SE-like cell line is also mentioned in the literature, JKT-1, 23 it was later shown to not express traditional SE markers and therefore is not considered representative of SE. 24 Additionally, there is an interest in the field to study the TE phenotype, for its different biology and because of the distinct clinical approach in retroperitoneal metastases, but TE is underrepresented in available TGCT cell lines, since it is not easy to culture and to immortalize as cell lines (Table 1).
List of available TGCT cell lines and their respective histological characteristics.
CH, choriocarcinoma; EC, embryonal carcinoma; SE, seminomas; TE, teratoma; TGCT, testicular germ cell tumors; YST, yolk sac tumor.
Drug toxicity in healthy cells and tissues is very important to test. In this sense, cell lines that are representative of nonneoplastic tissues also are key for drug testing studies, as they allow a first screening regarding toxicity and may aid in unveiling drug selectivity for cancer cells before the beginning of animal studies. In the specific case of testicular cancer, a Sertoli cell line (FS1) established from testicular tissue of a patient with Frasier syndrome has been used in the past as a control for testicular microenvironment in testicular cancer drug experiments. 45 MPAF, which is a human adult fibroblast cell line, and Hs1.Tes, comprising fibroblasts isolated from the testis of a Caucasian male, have also been used for this purpose. 45 Although this strategy represents an approximation and attempt to verify off-target effects of drug compounds, such immortalized cell lines are not completely “normal” cells, as they go through a genetic modification process so they can divide indefinitely and their phenotype can change with long-term culture. 46 Therefore, all these assays need to be validated with more complex in vivo experiments later in the process before moving to clinical trials.
In order to further improve in vitro models, the tumor microenvironment (TME) component is also important and has been shown to have considerable influence in TGCT. For that, co-cultures of tumor cells with other cell types that are extensively present in the TME will improve the reliability of these models, so that researchers have further knowledge on treatment response in the tumor and its surrounding cells. TGCT tumor communication with the TME has been studied with co-cultures,47–50 and the tumor relationship with fibroblasts/endothelial cells was demonstrated to influence microRNA levels, specifically the miR-371-373 cluster, 47 which has a central role in these tumors. 51 In another study, in which the intrinsic functions of cancer-associated fibroblasts (CAF) in GCT were investigated, it was shown that SE and EC activate CAFs, shaping their interaction with macrophages and subsequent recruitment of immune cells to the TME. 52 Regarding immune cell interaction in TGCT, activation markers and pro-inflammatory cytokines were altered in a co-culture of the TCam-2 SE-like cell line with monocytes or T cells, with more influence on IL-2, IL-6, and TNFα, showing the importance of the immunological TME in this TGCT subtype specifically.48,50 On the other hand, NTERA-2 cells were unable to activate T cells and monocytes or induce the release of pro-inflammatory cytokines in a co-culture model, evidencing the distinct immunomodulatory properties between representative SE and NST cell lines. 49 These findings may be translatable, as from a clinical standpoint, it is known that SE presents a more inflamed TME, whereas the immune infiltrate of NST is usually less pronounced. Hence, using this type of co-culture approach in the testing of novel targeted therapies could give researchers additional hints on the complex interaction with the microenvironment, which may allow for adjusting next steps in more complex in vivo models.
GCTs are theorized to develop from primordial germ cells (PGCs) that fail to differentiate. 53 In order to effectively study these mechanisms, generation of primordial germ cell-like cells (PGCLC) is warranted. Importantly, considerable advances have been made in the last years regarding long-term maintenance of PGCLs. 54 A novel model that allows for a perpetual expansion of PGCLCs was reported, with the expansion of the cell population maintaining the germline features and the ability to convert them into embryonic germ-cell like cells. 55 Models such as this one are important to study the origins of GCT and to identify novel targets with therapeutic potential. Thus, although transformation of PGCs into GCTs has not been achieved, it is currently an active area of research and advances may be done in the coming years to further understand GCT tumorigenesis and etiology.
Another factor that is essential when testing novel therapies for cancer is the effect of the drug on tumor architecture. Conventional two-dimension (2D) in vitro cultures are not adequate models for cell–cell and cell–matrix interactions, and typically have unlimited access to oxygen, metabolites, and nutrients, not properly mimicking the in vivo scenario. 56 Thus, as all tissues in the body have three-dimensional (3D) conformation, spheroid models have been used in the last decades to mimic better tissue microstructure. There are currently several different types of spheroid models for cancer cells, including hanging drops, ultralow attachment plates (ULA), and matrix-based models, each with its own advantages and disadvantages for disease modeling and therapy testing. 56 In 2022, a SEM-1 cell line spheroid model was established in order to study the effect of metformin, which prevented the 3D spheroid formation and blocked cell invasion and migration. 57 Furthermore, TCam-2 SE-like cell line was also used for spheroid formation in ULA plates, in order to study E-cadherin expression regulation. 58 Hence, more studies employing this type of model are expected with other TGCT cell lines, as they are being increasingly used in cancer research. However, spheroids from immortalized cell lines also have disadvantages, as they still do not accurately capture tumor phenotypic diversity, and show proteomic/transcriptional changes after a high number of passages. 56 In order to avoid these disadvantages, patient-derived spheroids have been used in recent years. Nevertheless, there is only one published study on the successful establishment of patient-derived TGCT spheroids, showing that stemness- and hypoxia-related genes were highly expressed in GCT spheroids when compared with monolayer cultures. 59 Thus, this is a type of model that needs to be further explored, as it may have potential for a closer to the clinic approach of therapy testing.
In a different approach to 3D culture, organoids have also been in the spotlight for the last few years. Unlike spheroids, which are very time-limited in their use, organoids can be maintained for several months and are derived from either pluripotent, tissue-resident stem, or differentiated cells from healthy or unhealthy tissues, such as tumors. 60 Despite the advantages, organoids are much more challenging to generate and demand more specific skills and supplies to maintain them. 61 Although there is published literature on testicular organoids to study the germ cell niche and drug reproductive toxicity, 62 there are not yet any published articles of organoid models for TGCT. This makes it a promising field to study in the near future, as these models use patient-derived material, are closer to the actual individual patient tumor biology for testing novel tailored therapies when compared to simple 3D spheroids and 2D in vitro models, and have been used in the past for screening for targeted therapies and as prediction models of response to novel drugs. However, the high variability and the complex mixed histology of most of TGCTs, with different proportions of several histologic subtypes, are significant hurdles to establish and maintain accurate patient-derived spheroids and organoids from these tumors, representing a significant gap in the field, given the success of such models in other somatic tumor types. In general, the lack of success in generating such models is a significant limitation in TGCT preclinical research, and the successful development of patient-derived organoids would bring valuable knowledge on TME research, tumor-stroma interactions and tumor heterogeneity, allowing targeted therapies to be more accurately tested before passing to clinical studies.
Clinical importance of in vitro models
Overall, there are a variety of in vitro models that can be applied in TGCT preclinical studies. An extensive number of cell lines has been described and used to study various parameters of TGCTs, both in monolayer and 3D, for drug testing and resistance mechanisms experiments. However, more cell lines representative of SE and TE are warranted to properly study the molecular and biological pathways of these tumor subtypes and avoid a bias with only EC, which comprises most of the available cell lines. Additionally, areas of future improvement include the generation of 3D patient-derived spheroids and organoids, which have not been extensively studied for this tumor model. This can only happen with biologists, clinicians, and pathologists working together in multidisciplinary teams, in order to develop proper cellular models derived from patients that have to be well characterized and followed in a clinical point of view. Such models are essential for specific targeted therapies and novel biomarkers to eventually go for clinical studies and have real impact in TGCT patient management.
In vivo models
Mouse models are a powerful tool to study the biological processes that result in the initiation and progression of primary and metastatic neoplasms and can provide a platform to develop novel therapeutic strategies for treating patients. The first mouse TGCT model was established when Stevens and Little discovered that inbred 129/Sv mice spontaneously develops testicular TE and teratocarcinomas. 63 Since then, multiple additional TGCT models have been developed in mice and other species (Figure 1).
In vivo TGCT models. (a) TGCTs can arise in mice through spontaneous tumor formation or can be induced experimentally via gonad transplants, chromosome substitutions, or genetic mutations. (b) TGCTs can arise in other nonrodent animals, including genetically engineered zebrafish and pigs, or spontaneously in dogs, horses, and pigs. (c) TGCTs can also be modeled in vivo through xenografting of established cell lines or patient samples into different mouse sites or in the chick chorioallantoic membrane.
Spontaneous testicular germ cell tumors in mice
Stevens established the first in vivo TGCT model when he discovered that the 129/Sv inbred mouse strain spontaneously develops TGCTs, albeit rarely, with 1% of male mice developing TGCTs before 20 days of age. 63 These tumors typically arise unilaterally, with greater frequency in the left testis, and consist of TE that contain differentiated tissue including neural and epithelial cells, as well as cartilage and bone.63,64 Tumors collected from younger mice also contain undifferentiated EC cells, classifying them as teratocarcinomas. These mice have played an important role in understanding TGCTs, including determining its prenatal origins and establishing the importance of the genetic background in tumor development. However, the extremely low tumor incidence limits its utility for researchers.
Different mutations introduced into 129/Sv mice were identified to increase TGCT incidence, including spontaneous and genetically engineered mutations described here and below, and thoroughly reviewed previously. 65 A spontaneous mutation in Steel was found to influence germ cell development; germ cells were reduced in heterozygous Steel mutant mice and completely depleted in homozygous Steel mutant mice. 66 When backcrossed onto the 129/Sv background, the Steel mutation modestly increased the incidence of testicular tumors to 9%. 67 It subsequently was found that the Steel mutation deletes KIT ligand (KITL), which is commonly mutated in patient TGCTs and implicated in TGCT development. 67
Another mutation, Ter, significantly increases the incidence of testicular TE. 129/Sv-Ter/+ mutants have a 20% tumor incidence which increased to 90% in 129/Sv-Ter/Ter mutant mice, including an increase in the incidence of bilateral TGCTs. 68 These mice also have smaller testes and show germ cell depletion starting as early as E15. Interestingly, mice on the C57BL/6 background with the Ter mutation also show germ cell depletion but not predisposition to testicular TE. This indicates that germ cell deficiency alone is not sufficient to induce TE formation and that 129/Sv mice are distinct in their susceptibility to TE formation. 68 The Ter mutation was identified as a loss-of-function mutation in Dead end 1 (Dnd1), which is essential for PGC migration and survival in mice and zebrafish.69–71 Germ cell depletion and spontaneous TGCT formation in Dnd1Ter/Ter mutant mice is partially related to BAX-mediated apoptosis. Dnd1Ter/Ter mice in a mixed background have germ cell depletion but no spontaneous TGCTs, while Bax knockout partially rescues the germ cell defect by 50% and results in TE formation. 69 Interestingly, the same Dnd1Ter/Ter Bax-/- genotype on a pure C57BL/6 mouse background is not associated with TGCT formation. 72 This suggests that the exceptional susceptibility of 129/Sv mice to spontaneous TGCTs may stem from a higher rate of transformation of mutant germ cells or a defect in mutant germ cell apoptosis compared to other mouse strains. Further studies comparing Dnd1 function in 129/Sv and C57BL/6 mice determined that Dnd1 promotes mitotic arrest in germ cells, with failure of germ cells in 129/Sv mice to arrest contributing to TGCT predisposition.72,73 The Ter mutation and subsequent identification of the Dnd1 gene have yielded important insights into roles for apoptotic pathways and mitotic arrest in germ cell development and TE formation. Despite its strong association with testicular TE in 129/Sv mice, Dnd1 mutations have not been linked to TGCT risk in human patients.
Germ cell tumors in chromosome substitution mouse strains
Risk loci for TGCTs have been identified by linkage analysis in which 129/Sv-Ter+/– mice were crossed with MOLF/Ei mice, an inbred mouse strain derived from Mus musculus molossinus that is not susceptible to spontaneous TGCT. 74 The results indicate that MOLF/Ei-derived loci can modulate the susceptibility of 129/Sv mice to testicular development, with several tumor-modifying loci identified on chromosome 18 and 19.75,76 As a result, the 129.MOLF-Chr 19 consomic strain was derived, in which chromosome 19 from 129/Sv mice is replaced with the homologous MOLF/Ei chromosome 19. 74 These mice have a high incidence of testicular TE with an increased proportion of bilateral tumors, phenotypes linked to five regions on the MOLF-derived chromosome 19, which contains genes known to be associated with TGCTs. 77 Interestingly, the 129-Chr18MOLF consomic strain, in which 129/Sv mice have a MOLF/Ei-derived chromosome 18, is resistant to TGCT formation, suggesting that loci on the MOLF-derived chromosome 18 may have a protective effect against tumor development when on a 129/Sv genetic background. 78 The 129.MOLF-Chr19 mouse line makes it possible to study complex TGCT-associated traits on the MOLF-derived chromosome 19 and identify novel genes and gene interactions involved in testicular TE development. 74 These mice are fertile, making them an easier model to maintain and leading to a number of elegant studies highlighting how failure of cell cycle arrest and sex specific differentiation as well as retention of pluripotency, contribute to TGCT formation. 79 These chromosome substitution mouse strains highlight the impact of genetic background on TGCT formation and the complex multigenic nature of tumor susceptibility.
Germ cell tumors from genital ridge engraftments
Experimental production of TGCTs by tissue engraftment has been used to study early TE formation and the interaction between TE development and host testis tissue. 80 Testicular TE can be experimentally produced in mice by implanting the genital ridge from E12.5 to 13.5 embryos onto adult testes or spleen. The location of the graft, the age of the genital ridge, and the genetic background of the genital ridge graft and host mice all play a factor in TE incidence. TE arising from grafts onto the testis are larger and more histologically complex compared to those from grafts onto the spleen. A higher tumor incidence is also observed with genital ridge grafts from E12.5 embryos compared to E13.5 embryos and when the donor genital ridge or the adult host testes is derived from 129/Sv mice. 81 By contrast, genital ridge grafts from the LTXBJ (LT) mouse strain background onto 129/Sv host testes do not develop germ cell tumors. 82
Linkage analysis using simple sequence length polymorphisms identified a risk locus on chromosome 18, named ett1 (experimental testicular tumor 1), that confers susceptibility for engrafted TGCTs. 82 Genital ridge grafts from a congenic strain containing the 129/Sv ett1 locus on the LT background developed TE when implanted into host mice on a mixed background. By contrast, grafts from WT LT mice implanted onto the same mixed host background did not form TE. 82 One gene within the ett1 locus is Mc4r (melanocortin four receptor), which contained a 129/Sv-specific missense mutation. Interestingly, double Mc4r-/- Ter-/- mutants on the resistant LT strain background showed a high incidence of ovarian tumors but no TE development. 83 However, the 129/Sv-specific point mutation in Mc4r is sufficient to establish engrafted TEs; genital ridge grafts from LT mice containing the Mc4r point mutation implanted onto LT host mice developed testicular TE that contained tissue from one to three germ layers. Conversely, genital ridge transplants from LT mice with wild-type Mc4r fetal testes onto host LT testes did not result in TE. 84 This would also suggest that the genotype of the donor fetal tissue is more important in TE formation compared to the host genotype. In sum, tissue engraftment TGCT models provide a unique avenue, in a specific developmental and experimental context, to study the influence of the host and donor genetic background on TE formation.
Genetically engineered mouse teratoma models
The role of different genes in TGCT development can also be studied by generating genetically engineered mouse mutants. Pten is a tumor suppressor gene that is essential for normal development; Pten-/- mice die during early embryonic development and Pten+/- mice have a high rate of spontaneous tumor formation in a wide range of tissues. 85 In humans, Pten is expressed in normal germ cells and in germ cell neoplasia in situ, but its expression is reduced in TGCT suggesting that Pten loss may promote TGCT progression. 86 Homozygous Pten deletion in PGCs at E8.5, using a TNAP-Cre driver, resulted in dedifferentiation of PGCs, partially through the PI3K/AKT signaling axis, and bilateral TGCTs with differentiated and undifferentiated components detected at P0. 87
GWAS studies identified multiple independent risk loci associated with TGCTs located in Dmrt1 (double sex and mab-3 related transcription factor 1).88,89 Dmrt1 is a widely conserved transcription factor expressed specifically in the gonads and required for testis differentiation. Additionally, Dmrt1 acts as a dose-dependent suppressor of TGCT in mice. Dmrt1–/– mice have a 90% incidence of testicular TE as compared to 4% in Dmrt1+/– mice. 90
GWAS studies also identified TGCT-associated SNPs located in Dazl (deleted in azoospermia-like), which encodes a conserved RNA-binding protein that promotes PGC differentiation by targeting pluripotency genes.91,92 Dazl loss in mice causes failure of germ cell commitment, eventually resulting in apoptosis and germ cell depletion. 93 In 129/Sv mice, Dazl loss increases the incidence of testicular TE to approximately 30%, with tumors detectable as early as 4 weeks of age. 93 As in Ter mutant mice, inhibiting germ cell apoptosis with heterozygous Bax deletion increases the rate of spontaneous testicular tumors, suggesting that mutant germ cells that fail to undergo germline commitment typically undergo Bax-mediated apoptosis to prevent testicular TE. 93 These genetically engineered mice are important models of human TGCTs as they target genetic modifications that are conserved in human patients. They also provide valuable tools for studying Type 1 testicular TE but are unable to reproduce the Type 2 malignancies that are common in human patients.
Genetically engineered malignant germ cell tumor models
Malignant TGCTs have been successfully modeled in mice via Pten inactivation and constitutive Kras activation specifically in germ cells around E12.5. 94 Termed gPAK mice (germ-cell specific Pten and Kras mutant mice), these mice show a TGCT incidence of 75% and develop mixed germ cell tumors with differentiated TE tissue and populations of pluripotent EC cells. 95 These tumors also frequently metastasize. Like in humans, gPAK TGCTs are highly chemosensitive, with a decrease in tumor volume and increase in survival for mice treated with the standard of care chemotherapy regimen for humans.95,96 This provides an immunocompetent mouse model that features the spontaneous development of malignant, metastatic TGCTs that originate specifically in embryonic germ cells, and are extremely sensitive to chemotherapy treatment. However, gPAK mice are unable to fully recapitulate human TGCTs. In patients, TGCTs originate from a germ cell neoplasia in situ precursor lesion, are typically aneuploid, and manifest as SE in more than half of cases. In contrast, the precursor lesion for gPAK mouse tumors is unknown, the neoplasms are diploid, and gPAK mice never develop SE. These limitations may suggest that there are other factors in TGCT development that are not represented in the gPAK model or that the differences in mouse and human development limits the TGCT subtypes that can arise from a mouse model.
Spontaneous germ cell tumors in other mammals
Other nonmodel mammals can be used to provide insight into germ cell tumors. Rats spontaneously develop testicular tumors, but unlike mice, they typically do not develop germ cell tumors but instead develop interstitial (Leydig) cell tumors, serving as a model for human interstitial cell tumors. 97 Dogs and horses have also been reported to spontaneously develop SE and TE, with a portion of them diagnosed in cryptorchid testes, potentially linking cryptorchidism to testicular tumor incidence as in humans.98,99 Pigs have been used to confirm the conservation of germ cell commitment mechanisms; a small percentage of boars spontaneously develop seminomas, genetically engineered Dazl-deficient boars lack germ cells at 11 weeks of age, and Dazl-deficient female pigs develop ovarian TE at a high rate.93,100 These animals may provide alternative systems for studying the development of germ cell tumors and potential therapeutic treatments.
Nonmammalian germ cell tumor models
Zebrafish provide an additional vertebrate system to study human germ cell tumors. Germline development is broadly conserved between zebrafish and humans; PGCs in both zebrafish and humans migrate during embryogenesis and aberrant germ cell differentiation results in germ cell tumors. 101 Zebrafish were found to spontaneously develop germ cell tumors at a low rate between one and two years of age. 102 A forward genetic screen using random ENU mutagenesis to find genomic instability mutants found one mutant male line that had a 6% incidence of invasive germ cell tumor and 53% of noninvasive SE compared to control males. 103
A mutant line with a Lamc1cz61 mutation, named tgct, was established as a heritable germ cell tumor model and has a 30%–50% incidence of spontaneous germ cell tumors in zebrafish older than 18 months. 104 TGCT mutants have a mutated Alk6b, encoding a bone morphogenetic protein receptor, which disrupts the BMP pathway required for germ cell differentiation. The tumors are histologically similar to SE, contain undifferentiated germ cells, have impaired spermatogonia differentiation, and are sensitive to chemotherapy treatment. 104 Transcriptomic analysis revealed that the zebrafish tumors have a gene expression signature that reflects a combination of EC, SE, and YST features. 105 In human TGCTs, perturbed BMP signaling correlates with impaired germ cell differentiation; 60%–90% of differentiated TE had intact BMP signaling compared to only 10%–22% of undifferentiated EC and SE. 106 Another zebrafish germ cell tumor model in which over 90% of zebrafish between two to three years of life develop tumors, many of them TGCTs, features a loss-of-function mutation in Lrrc50 (leucine-rich repeat containing protein 50). 107 Histologically, these tumors resemble SE: they are morphologically uniform, predominantly contain early germ cells, and have strong expression of Ziwi, the fish ortholog of the human SE marker. 107 In humans, Lrrc50 loss-of-function mutations are found in SE, indicating a potential conserved role in SE development. 107
Two transgenic fish models have also been established for TGCTs: one containing a simian virus 40 T-antigen (TAg) transgene with the fugu lymphocyte-specific protein tyrosine kinase promoter (flck) (Tg(flck:TAg)) and one containing the stem cell leukemia (scl) and the lim domain only one transgene (lmo1) driven by flck (Tg(flck:scl/lmo1)). 108 A total of 16%–20% of Tg(flck:TAg) transgenic offspring develop germ cell tumors, compared to <4% of wild-type fish, with disorganized testicular structures that contained primarily spermatogonia and spermatocytes. Subclinically, 90% of Tg(flck:TAg) transgenic fish over 36 months of age were found to have enlarged testes that comprised primarily of spermatogonial-like cells compared to 50% of wild-type fish. 108 These mutant zebrafish models of germ cell tumors provide a unique advantage in modeling SE, a common TGCT subtype that is represented in over half of TGCT patients yet has been difficult to establish in mice.
Germ cell tumor cell transplant models
Genetically engineered mouse models are an invaluable tool for studying TGCTs, but existing models are limiting as they do not represent all tumor subtypes and have certain genetic features that are not fully representative of human TGCTs. To address this, researchers use transplant systems with established human TGCT cell lines and patient tumors specimens that allow for comprehensive in vivo studies.
The chorioallantoic membrane (CAM) assay is a transplant model in which tumor cells mixed with Matrigel are transplanted onto the vascularized membrane of chick embryos. This system can be used to study antitumor drugs and cancer hallmarks including angiogenesis, growth, and metastasis. 109 Using this assay, anti-angiogenic compounds HP-2 and HP-14 were found to decrease cell growth in cisplatin-sensitive and cisplatin-resistant 2102EP tumors 110 ; the chimeric inhibitor animacroxam decreased metabolic activity in 2102EP tumors 111 ; histone deacetylases inhibitors increased apoptosis and cell cycle arrest in NCCIT and 2102EP tumors 112 ; and knockdown of VIRMA, a component of the methyltransferase complex, decreased tumor size and increased sensitivity to cisplatin-treatment in NCCIT tumors. 113 These studies show the utility of the CAM assay in developing novel therapeutics for the treatment of TGCT.
TGCTs can also be studied with cell-derived xenografts (CDX), which are tumors grown from cultured TGCT cells injected into mice. This provides a reliable, reproducible, and easily scalable way to study specific TGCT subtypes that do not exist in other mouse models, including YST, SE, and cisplatin-resistant tumors, as well as responses to cancer therapeutics (Table 2).
Outline of preclinical models commonly used in TGCT drug research.
CAM, chorioallantoic membrane; CH, choriocarcinoma; EC, embryonal carcinoma; SE, seminomas; TE, teratoma; TGCT, testicular germ cell tumors; YST, yolk sac tumor.
Early grafting studies established the tumor-propagating potential of murine EC cells by injecting dissociated tumor cells into 129 mice via intraperitoneal or subcutaneous routess. 114 This resulted in the growth of teratocarcinomas, showing that EC cells are multipotent with differing capacities for differentiation. 114 Numerous xenograft models with established human TGCT cell lines implanted into mice also have been characterized. For example, 1411H and 1411HRQmet cells establish mixed EC and YST neoplasms 115 ; TCam-2 cells give rise to tumors with SE-like histology22,31; NTera-2 cells implanted into the mouse testes develop teratocarcinomas with EC and evidence of neuroepithelial and glial differentiation. 116
These xenograft models are used to study novel therapeutic treatments. Tumors established from transplanted TCam-2 or NTera2 cells had a lower tumor volume and growth when treated with a combination of cisplatin and inhibited RANKL inhibitor, a member of the TNF (tumor necrosis factor) superfamily, providing a potential therapeutic alternative for lower cisplatin dosages with combination therapy. 117 Another alternative therapeutic treatment involves specific targeting of pluripotent TGCT cells with differentiating agents. Tumors established from NT2D1 cells treated with thioridazine showed delayed growth in mice, highlighting the therapeutic potential of differentiation treatment of pluripotent TGCT cells. 96 Inhibition of PIKfyve, a phosphatidylinositol kinase involved in lysosome homeostasis, was found to selectively induce apoptosis in pluripotent TGCT cells; Tera-2, Ntera22 and GCT27D.C1 cells treated with WX8, a PIKfyve inhibitor, had increased cell death, which was suppressed upon cell differentiation. 118 WX8 treatment of subcutaneous Ntera-2 and GCT27D.C1 xenografts decreased tumor initiation and growth with reduced population of pluripotent cells compared to vehicle controls. Inhibition of Pin1, a peptidyl-prolyl cis/trans isomerase that is correlated with poor TGCT patient prognosis, has also been found to decrease proliferation, increase apoptosis, downregulate pluripotency genes, and decrease tumor size in P19 and NCCIT transplant models. 119
Many studies have also focused specifically on targeting cisplatin-resistant tumors, as cisplatin resistance drastically decreases the prognosis for patients. TEX11, an X-linked meiosis-related gene, was found to be upregulated in cisplatin-resistant cells, and silencing TEX11 in cisplatin-resistant and cisplatin-sensitive xenograft tumors repressed tumor growth. 120 Cisplatin-resistant tumors have also been found to be highly sensitive to DNA methylation inhibitors. Cisplatin-resistant tumors treated with one dose of cisplatin and either GSK-J4, 121 a histone demethylase inhibitor, or guadecitabine, 122 a DNA methylation inhibitor, resulted in tumor regression in mice. This suggests that targeting methylation pathways in combination with cisplatin may be a more effective treatment and allow for lower cisplatin doses.
Increasingly, the importance of the tumor microenvironment in TGCT growth has been established. NCCIT and NT2 xenografts with miR-125b microRNA over-expression had smaller tumors compared to xenografts with repressed miR-125b expression. 123 miR-125b was found to influence the tumor microenvironment by decreasing tumor-associated macrophage recruitment by regulating chemokines CSF1 and CX3CL1, resulting in a decreased tumor size.
CDX models have provided many essential insights for TGCT growth and developing therapeutics but have limitations in modeling the heterogeneity of human cancers. They are created with homogeneous populations of cells that are adapted to cell culture conditions and injected, typically, into ectopic sites of immune-compromised hosts, providing an incomplete representation of human tumors and their microenvironment. To address some of these issues, researchers use patient-derived xenografts (PDX), in which patient tumor tissues are directly implanted into immunodeficient mice. This can preserve the diversity and histological architecture of patient tumors and accurately recapitulates the histology, genetic characteristics, and treatment response of the parental patient tumor.
The development of PDX for GCT has lagged compared to that of other more common malignancies. However, PDX models have been established in mice with (1) orthotopic implantation onto the testes, resulting in models of YST (TGT1), EC (TGT12), CHC (TGT17, TGT38), and a cisplatin-resistant YST from a refractory tumor (TGT44)124–126; (2) subcutaneous implantations into the flank including a mixed yolk sac and immature TE (TC1), a cisplatin-resistant YST derived from a resistant patient tumor (TC4), and a mixed YST and TE tumor (TC5)127,128; and (3) orthotopic intercranial injections to model metastatic intercranial tumors, which preserve mutations or overexpression of KIT, which can result in uncontrolled proliferation and resistance to apoptosis. 129 Cisplatin-resistant PDX models have also been established by treating cisplatin-sensitive PDX mice with increasing dosages of cisplatin, resulting in matched cisplatin-sensitive and resistant PDX models. This includes studies with YST (TGT1X and TGT1XR), EC tumors (TGT12X and TGT12XR; TGT34X and TGT34XR), CHC (TGT38X and TGT38R), and mixed tumors containing YST, EC, and CHC (TGT21X and TGT21XR). 130 These tumor xenografts largely resemble their parental tumors in terms of histopathology, genetic variation, and response to chemotherapy, even after multiple in vivo passages in mice, providing a valuable tool to accurately model patient tumors. However, there have been no successful PDX models of human SE reported to date. Human mixed TGCTs containing SE and NST components will grow as a pure NST in vivo.128,130 The absence of PDX models for pure SE remains a hurdle for in vivo research in TGCT as these represent a different biological pathway of tumor initiation when compared with EC, and may be due to several biological factors, including lower tumor aggressiveness and subsequent lower proliferative rate in vivo, or a dependency on TME factors that difficult cell survival.
PDX models have been used to test therapeutics in patient tumors. Tyrosine kinase inhibitors (TKI) are among the targeted therapies tested. Sunitinib and pazopanib inhibited angiogenesis in cisplatin-resistant and cisplatin-sensitive xenograft tumors, resulting in decreased tumor volume, and lapatinib, an ErbB1 and ErbB2 inhibitor, prevents aberrant activity of the tyrosine kinase receptors and reduces tumor growth.124–126 Inhibition of mTORC1/2 or MDM2 slowed tumor growth and increased sensitivity to cisplatin treatment, showing preclinical efficacy.127,128 These studies reflect how PDX are a valuable model to identify novel therapeutics and therapeutic targets for patient treatment, although the full therapeutic response is limited by the immunodeficient mouse host used for transplants.
Clinical importance of in vivo germ cell tumor models
In vivo GCT models are an invaluable tool to identify biological processes implicated in tumor susceptibility, initiation, and progression. These tumor models have identified or validated several genetic alterations that contribute to TGCT development, allowing for an increasingly comprehensive list of TGCT risk factors in men. They also can provide a model for identifying and validating clinical TGCT biomarkers and therapeutics. Recently, serum from the gPAK mouse model, which develops malignant, metastatic TGCTs, was found to have elevated levels of microRNAs from the mmu-miR-290-295 cluster, which is orthologous to the human hsa-miR-371-373 cluster. A microRNA from this cluster, miR-371a-3p, has emerged as a useful biomarker for malignant TGCT.131,132 This demonstrates the clinical importance of in vivo GCT models in determining risk of TGCT incidence, diagnosing patients with improved clinical biomarkers, and developing novel therapeutics for improved patient outcomes. Hence, development of animal models that allow to fill gaps in GCT biology knowledge can definitely aid in testing novel targeted therapeutics, to make sure they are safe and effective for TGCT patient treatment, and to take them into clinical trials.
Clinical applications
As discussed previously, TGCTs are hallmarked by high sensitivity to platinum-based chemotherapy. However, there is a small percentage of patients (about 20% of metastatic cases) that develop resistance to the treatment, frequently resulting in patient death. 1 The current care options for patients with refractory platinum-resistant TGCT are widely discussed in the European Society for Medical Oncology (ESMO), European Association of Urology (EAU) and American Society of Clinical Oncology (ASCO) guidelines.133,134
Relapsed TGCT patients can be treated either with conventional-dose chemotherapy (CDCT) or high-dose chemotherapy (HCDT) with peripheral-blood stem-cell transplant (PBSCT), with both strategies having curative potential. 135 There has been a push to find prognostic factors that may aid in the treatment of this patient subset. Histology, primary tumor location, treatment response, progression-free interval after first-line treatment, levels of alpha fetoprotein, human chorionic gonadotrophin, and the presence of liver, bone, or brain metastases at salvage therapy time have been set in the past to build a prognostic model to guide salvage chemotherapy strategies. 136 CDCT or HCDT with PBSCT may cure 20%–60% of patients in salvage chemotherapy, depending on patient relapse and resistance status. 137
Several clinical studies have been performed in this setting, such as a multicentric phase-II trial using TIP chemotherapy (paclitaxel, ifosfamide and cisplatin) as salvage for patients with metastatic GCT who have failed initial BEP (bleomycin, etoposide, and cisplatin) chemotherapy. This trial has concluded that a substantial proportion of patients, especially the ones with good prognosis, can achieve long-term failure-free survival. 138 Another study using this chemotherapy combination also achieved a durable cure rate (63%) in a high proportion of relapsed patients in the trial (70% of 46 patients). 139
To date, the efficacy of CDCT and HCDT with PBSCT overlaps significantly, and therefore there is still controversy around which regimen is best for treating these patients. 137 It is important to perform these types of treatment courses in reference centers with expertise in GCTs, so that patients are followed by multidisciplinary teams of GCT specialists and so that large cohorts are gathered to strengthen clinical data. 137
In the last decades, personalized medicine in cancer has emerged as a method for tumor treatment considering specific characteristics of each individual patient, such as gene expression changes, mutations and immune TME. The concept of targeted therapy appeared in the scope of that approach and can be defined as any agent that is more effective in tumors with specific molecular alterations, depending on the given agent. 140 These include angiogenesis inhibitors, immunotherapy, PARP inhibitors, TKI and epigenetic compounds, among others. Clinical trial studies performed in TGCT patients with targeted therapies of different classes and their respective results are summarized in Table 3.
Summary of the different clinical trials performed with targeted therapies in TGCT.
Cisplatin acts by causing DNA damage through the formation of intra- and inter-strand cross-links, subsequent platinum-DNA adducts formation, and eventually, if these cross-links are not repaired, DNA double-strand breaks are induced. 161 The homologous combination repair (HR) pathway is responsible for repairing double-strand breaks in DNA and is hampered in some TGCT cell lines. 162 Given this information and the importance of the cross-links in the cisplatin mechanism of action, it can be assumed that the high sensitivity of TGCT to cisplatin may be in part from impairment in some of the DNA repair mechanisms. In this regard, tumors that have shown downregulation in the HR pathway, poly (ADP-ribose) polymerase (PARP) inhibitors have shown promise, 163 including significant therapeutic effect in preclinical studies with TGCT models, both as single therapy and in combination with cisplatin, potentiating its effect.164–166 Thus, this characteristic of TGCTs may represent a potential target for platinum-refractory tumors. However, in a proof-of-principle phase-II clinical trial with relapsed/refractory metastatic germ-cell cancer olaparib demonstrated no activity in heavily pretreated GCT patients as a single agent, with only one patient presenting stable disease (SD). 147 Along with those, further clinical studies were performed with other compounds for relapsed patients and did not achieve clinically significant results for TGCT.146,148 Reasons for this may include the fact that patients included in the clinical trials were not generally selected based on homologous recombination pathway status, raising the possibility that the lack of success in these trials may be due to early study design features rather than the underlying biological premise. The one patient that presented with SD after olaparib treatment in the phase-II trial was remarkably the only patient that presented a BRCA mutation, 147 thus showing impairment in the HR pathway. Thus, in future trials with PARP inhibitors in TGCT, pre-enrollment assessment of HR signatures or BRCA mutations testing could be performed in patients, in order to aid in candidate selection and to improve therapy success.
Anti PD-L1/anti PD-1 immunotherapy has been showing promise recently in the treatment of various solid tumors.167,168 The TGCT immune microenvironment is very diverse, with immune cell infiltration being very heterogeneous and common in this type of cancer, particularly in SE. 169 The immunoexpression of PD-1 and PD-L1 has prognostic significance in TGCTs.170,171 However, when translating to clinical studies, anti-PD-L1 did not achieve significant clinical efficacy in the trials performed so far,145,149,150 two of them with pembrolizumab, possibly due to the lack of expression (or low expression) of the immune checkpoint molecules in an unselected population (i.e., independent of immune checkpoints expression). Also, the complex tumor microenvironment of TGCTs raises the hypothesis that additional interactions need to be considered for immunotherapies to be clinically effective for TGCT patients. Thus, further research is needed on immunotherapy specific biomarkers for TGCT, on anti PD-L1 drugs (solo or in combination) and perhaps with other immune checkpoint inhibitors to check their therapeutic efficacy. On another perspective of immunotherapy, chimeric antigen receptor (CAR-)T cells are genetically engineered T cells, whose receptors are a combination of an endodomain, an anchoring transmembrane domain, and an ectodomain. 172 This therapy is engineered from cells from the patients themselves and has had remarkable success with hematologic malignancies in the past years. These are also being tested in solid tumors, although still with significant challenges. 173 A preclinical study in EC cell lines concluded that CD30-redirected CAR-T cells induced antitumor activity in vitro against three human EC cell lines, and also in vivo with a xenograft model. 174 Also, a phase-I clinical trial with 13 CLDN6-positive relapsed/refractory testicular cancer patients was performed using CAR-T cells targeting CLDN6 and showed patient safety and tolerability as well as promising clinical results, achieving 1 CR, 6 PR, 7 SD, and 5 PD, with an ORR of 33% and a disease control rate of 67%. 155 CLDN6 is expressed in the vast majority of TGCTs and has very infrequent expression in non-GCT cancers. 175 Thus, early-phase clinical trials with CAR-T cells have shown promising clinical activity in cisplatin-refractory TGCTs when compared with immune checkpoint blockade therapy. Despite this, further investigation is warranted in clinical trials with higher number of patients. Furthermore, comparative efficacy of the two therapies remains unknown due to a lack of head-to-head studies.
TKI are compounds that disrupt the pathways that lead to malignant cell growth. In TGCTs, despite showing in vitro activity in a cisplatin-resistant TGCT cell line, the TKI sunitinib showed modest activity (about 13% total response rate) in a trial with 33 cisplatin-refractory/multiply relapsed GCT patients. 141 Additionally, another phase-II trial also failed to confirm the activity sunitinib, despite having shown activity previously in the preclinical setting. 176 In another study, treatment of 6 patients with imatinib resulted only in PD for 5/6 patients and the other one maintained SD before progressing after 3 months. 142 Hence, the TKIs tested so far clinically for TGCT have shown poor activity in refractory disease patients. This might happen due to the low mutation burden of these patients for these pathways, and a rigorous biomarker-driven selection of TGCT patients for these clinical trials is warranted, in order to try and maximize the therapeutic efficacy of these targeted therapies.
The fact that TGCTs have a low mutation rate suggests that they may have a more epigenetic-centric mechanism in the basis of their pathogenesis.177,178 GCNIS and SE display global hypomethylation, whereas TE show methylation levels comparable with somatic tumors and EC possesses an intermediate level of CpG methylation.178,179 DNA methyltransferase inhibitors (DNMTi) have shown efficacy in preclinical models of TGCT in the past, both in parental and cisplatin-resistant cell lines. 180 The same was observed with histone deacetylase inhibitors (HDACi). 181 Some of these compounds, including decitabine and panobinostat, are already approved by the FDA and EMA for the treatment of hematolymphoid malignancies. 182 However, the results of the epigenetic drugs in clinical studies with TGCT patients were not very convincing so far, specifically for advanced patients.156,157,159 Hydralazine in combination with valproate (two drugs repurposed for their secondary effects as DNMTi and HDACi) were tested in a phase-II clinical trial with a NST patient, which showed SD after the trial. 159 Furthermore, more recently, guadecitabine, a second-generation DNMTi, was tested in a phase-I trial with 14 metastatic GCT patients and showed tolerability, with an ORR of 23%, and three patients also had SD. 158 Nevertheless, epidrugs are still a promising approach in this type of tumors and epigenetic modifications should be studied more in depth due to TGCT-specific background. These epigenetic compounds have been getting more selective in recent years with the advent of targeted therapies and now there are also dual inhibitors for different types of epigenetic enzymes, which are promising strategies for the near future.
In general, clinical trials using targeted therapies have shown modest clinical results for the group of patients that have metastatic refractory and platinum-resistant disease. This might be due to the great histological and biological heterogeneity in these tumors, 183 which have low tumor mutational burden and few targetable alterations, 177 making it difficult to use tailored approaches for treatment. These clinical results are likewise supported by the fact that cisplatin resistance is a multifactorial process, and targeting a single pathway or biomarker may be insufficient to overcome the resistance.184,185 An additional barrier is that research on biomarker-oriented studies and development of targeted agents in this tumor model lacks support and attention from agencies and industry, namely because of the low incidence, and also because of the remarkable sensitivity to cisplatin, with the majority of patients being cured of disease with standard chemotherapy regimens. 1 However, considering the tremendous impact of cisplatin exposure on these young males’ quality of life, and recent evidence stemming from testicular cancer survivors’ cohorts showing lower overall-survival and increased morbidity (by multiple factors, including second cancers. . .), 9 there is an urgent need to develop less toxic and more targeted therapies, bringing functional precision medicine based on proper thorough preclinical model investigation to TGCT patients. Currently, following a search for “testicular cancer” in clinicaltrials.gov, there are 88 clinical trials that include testicular cancer patients and are active and recruiting (data from clinicaltrials.gov as of 21st August 2025), with only six of those focusing on targeted therapies. We know from the results obtained so far that specific preclinical and clinical studies with novel targeted therapies are still warranted for the management of patients that relapse and develop resistance to platinum. Furthermore, these patients should be referred to TGCT reference centers for follow-up and treatment, so that they can be accompanied by GCT expert teams that can provide them with the best care possible. Aside from using novel targeted therapies, there have been clinical trials aiming to perform personalized treatment courses with conventional chemotherapy in combination with serum tumor marker measurement or imaging parameters, to either escalate or deescalate chemotherapy doses in patients with different prognosis.186,187 These approaches may optimize the management of TGCT treatment based on routine patient care.
Additionally, both the ESMO and the ASCO guidelines, as well as the EAU do not recommend novel targeted therapies for routine treatment of cisplatin-refractory patients yet, while encouraging that such patients should be included in clinical trials.133,134,188 They also highlight the lack of molecular testing of biomarkers for selecting inclusion in clinical trials as a possible reason for the lower success to date. Both classical, like serum tumor markers and imaging, and novel biomarkers, such as miR-371a-3p or CLDN6 testing, should be included in the criteria of enrollment to recruit patients for specific targeted therapy testing in clinical trials, so that clinicians have the correct data to provide patients with a truly tailored approach for TGCT treatment. The expression of miR-371a-3p reflects the natural course of the disease in TGCT, and it could be used as a good example of a biomarker with potential to use for treatment de-escalation, for instance. However, more biomarker driven clinical trials are still necessary for real-world setting validation.
Conclusion
Cisplatin resistance and over-exposure to cisplatin leading to toxicity remain important unresolved challenges in the TGCT field. Informative preclinical models are warranted to effectively tackle these disease settings. Available in vitro and in vivo models are the best current option for evaluating targeted therapies before these are ready to go into clinical studies. However, novel and improved preclinical models are required, so that researchers may have available tools that are closer to the individual tumor of a given patient, including its microenvironment and other complex interactions (Figure 2).
Overview of an appropriate preclinical/clinical pipeline with the objective to get approval for targeted therapies in TGCT.
Clinical trials in cisplatin-refractory TGCT patients focusing on targeted therapies, especially immunotherapy, DNA repair pathways inhibitors and epigenetic drugs have had modest results at most. A missing piece in this movement from preclinical models to actual clinical trials is the testing of actionable biomarkers in well-characterized patient samples included in clinical trials (with central histopathologic review 189 ), truly representative of metastatic and fully resistant disease, which has been the setting of focus on the available clinical trials. Furthermore, cisplatin resistance is widely known to be a multifactorial process, and most therapy options only target a single pathway, which could also be another reason for these results. For this, referral of TGCT patients to expert centers where clinicians work closely together with researchers is fundamental. The role of biobanks and pathologists in this translation from the lab bench to the bedside is important, and the only way precision medicine can be envisioned for TGCT patients.
