Abstract
Background:
Despite advancements in treatment of advanced testicular germ cell tumors (GCT), metastatic teratoma remains a significant challenge due to its intrinsic chemo- and radioresistance. Claudin-6 (CLDN6), a tumor-specific antigen, has shown promise in targeted therapies for various solid tumors, including metastatic GCT in early phase trials. Testicular GCT frequently exhibit a unique cytogenetic hallmark, isochromosome 12p (i12p), which is associated with tumor development and progression, underscoring its significance in oncogenesis and diagnosis, and potential as a potential biomarker.
Objectives:
We aimed to understand the expression of CLDN6 during progression from germ cell neoplasia in situ (GCNIS) to primary GCT and postchemotherapy teratoma, and to investigate persistence of i12p in postchemotherapy teratoma.
Design:
An observational cohort study.
Methods:
Using matched pre- and postchemotherapy formalin-fixed paraffin-embedded tumor tissue from patients with primary GCT and postchemotherapy teratoma identified from a national GCT registry, iTestis, CLDN6 expression, and presence of i12p were evaluated.
Results:
Twenty-two patients were eligible. The majority were diagnosed with mixed primary GCT (20/22, 91%) with teratoma the sole histologic component in 86% of postchemotherapy samples (19/22). CLDN6 expression was consistently observed in GCNIS (16/16, 100%) and primary nonteratoma GCT (20/20, 100%) though significantly lower in primary teratoma components (4/10, 40%) and metastatic teratomas (5/22, 23%). The mean proportion of CLDN6-positive cells was higher in primary nonteratoma components (87.5%) compared to primary (6.5%) and postchemotherapy teratoma (4.6%). i12p was also common in primary GCT (18/22, 82%), including pure teratoma (2/2, 100%), though concordance between i12p in primary orchidectomy and metastatic samples was lower (13/18, 72%).
Conclusion:
Our results demonstrate early overexpression of CLDN6 in GCNIS and high expression in nonteratoma components in primary GCTs, but limited expression in primary and posttreatment teratomas. CLDN6-targeted therapies may not be effective for teratoma. In contrast, i12p was prevalent across all stages of GCT progression, suggesting its potential as a biomarker for identifying viable GCT.
Plain language summary
Treatments targeting Claudin-6 appear very promising for advanced testicular cancer patients, as Claudin-6 is abundantly present on testicular cancer cells. However, some advanced testicular cancer transforms to teratoma, which is very difficult to treat. We explored whether teratoma also expresses Claudin-6. We also utilized Claudin-6 and another testicular cancer marker, isochromosome 12p, to characterize the evolution of testicular cancer from pre-cancerous to cancer to teratoma.
Background
Testicular germ cell tumor (GCT) is the most common cancer affecting young males aged 18–39 years. 1 Histologically, more than 90% of testicular tumors are GCTs, the vast majority of which are associated with the precursor lesion germ cell neoplasia in situ (GCNIS). Since the introduction of cisplatin-based chemotherapy in the 1970s, 2 testicular GCT is considered highly curable, with a 10-year relative survival approaching 95%. 3 Despite incremental improvements in survival with refinement of upfront and salvage options including high-dose chemotherapy (HDCT) and autologous stem cell transplantation, some patients still die from GCT. For decades, clinicians have focused their efforts on different chemotherapy combinations or earlier intensification of treatment in patients with high-risk disease to improve survival. While targeted therapies and immunotherapy using checkpoint inhibition have revolutionized the treatment of other cancers, they are largely unsuccessful in the management of GCT4,5 including teratoma, and there has been little focus on the development of novel therapies for this population.
Identification of Claudin-6 (CLDN6), a highly specific tumor-associated antigen, however, has led to a resurgence in new drug development focused on GCT. CLDN6 is a strictly oncofetal cell surface antigen without expression in normal human tissue, but high expression in many primary and metastatic solid tumors including GCT. 6 Early phase clinical trials of chimeric antigen receptor-T (CAR-T) therapies targeting CLDN6 with or without a CAR-T cell-amplifying RNA vaccine or antibody-drug conjugates have been undertaken and demonstrate promising results in individuals with treatment-refractory GCT,7–9 with studies ongoing.
Postpubertal teratoma poses a particular challenge for clinicians treating GCT. It is commonly noted as a residual mass or late relapse following definitive chemotherapy for advanced GCT and is largely considered chemo- and radiotherapy insensitive.10–12 The reasons for this are not well understood and surgical resection of teratoma remains the only reliable curative option, though it is not always feasible owing either to the anatomic location of the residual mass(es), or due to rapid tumor growth, referred to as growing teratoma syndrome, rendering it unresectable. In this context, metastatic unresectable teratoma remains an incurable disease.10–12
While the development of novel therapeutics for treatment-refractory GCT is expanding, few advances in clinical management of teratoma have been observed. High expression of CLDN6 in primary and metastatic GCT is well described,8,13 though its specific expression in teratoma, a histological subtype of testicular GCT, commonly observed following chemotherapy, is unknown. Given the early success of CLDN6 targeted therapy in relapsed refractory GCT, and sequential differentiation of teratoma from GCNIS and invasive GCT components, 14 it is plausible that CLDN6 expression may also be high in this entity. For the same reason, other biomarkers, including isochromosome 12p (i12p), a chromosomal aberration pathognomonic of invasive GCT, is also worthy of investigation as a biomarker for teratoma, though existing studies have yielded varying results.15–17
In this exploratory cohort of matched primary chemotherapy-naïve testicular GCT and postchemotherapy metastatic teratoma, we aimed to understand the presence of i12p and CLDN6 expression in nonteratoma and teratoma components and GCNIS, as a preliminary step toward novel drug development in this population. Moreover, we aim to understand the expression of i12p and CLDN6 during the progression from GCNIS to GCT and posttreatment teratoma.
Methods
Utilizing the national Australian testicular cancer registry, iTestis, a cohort of patients with testicular GCT comprising primary chemotherapy-naïve orchidectomy specimens (with and without teratoma component) and paired metastatic teratoma specimens from postchemotherapy resection was identified. Patient identification, sample collation, and distribution for analyses were conducted at the Walter & Eliza Hall Institute.
Participants were eligible if they had a clinical or histologic diagnosis of non-seminomatous germ cell tumor (NSGCT) including pure teratoma, had evidence of postpubertal teratoma in postchemotherapy resection of metastasis and had accessible archival tissue of both primary orchidectomy and postchemotherapy metastatic tissue. Archival tumor formalin-fixed paraffin-embedded (FFPE) specimens were retrieved and reviewed by a board-certified anatomical pathologist to identify areas of GCT and teratoma in the primary specimen, and, if present, viable GCT in representative postchemotherapy metastatic specimens. Clinical history was extracted from iTestis. There was no missing clinical data. One additional patient with GCNIS in the orchidectomy specimen that did not fulfill inclusion criteria for the matched pair analysis is presented for illustrative purposes only and was not included in the broader analysis.
Representative FFPE slides of each study participant were provided to BioNTech SE for CLDN6 analysis. Prior to CLDN6 immunohistochemistry (IHC), H&E slides were reevaluated at BioNTech SE by inhouse-pathologists for presence of nonteratoma and teratoma components, as well as GCNIS. Subsequent analyses were conducted separately in the identified components.
Immunohistochemical staining of FFPE sections was performed at the central histology laboratory of BioNTech SE, Mainz, Germany. Slides were manually stained with a mouse antihuman CLDN6 monoclonal antibody (clone 58-4B-2; CLAUDENTIFY®6 Kit; BioNTech Diagnostics, Mainz, Germany) according to the manufacturer’s instructions. In addition, each tissue was stained with a negative control reagent and each run included a slide containing embryonic rabbit kidney as a positive control to ensure consistent staining quality. H&E staining of consecutive tissue slides was performed for morphological correlation.
H&E- and CLDN6-stained slides as well as negative control reagent slides were digitized using a Hamamatsu NanoZoomer S360MD slide scanner, visually displayed using NDP.view2Plus (Hamamatsu, Japan) and scored by a board-certified anatomical pathologist. Scoring was based on both the percentage of stained cells and the intensity of staining: for each case, CLDN6 expression was estimated as the percentage of tumor negative for CLDN6 (score 0) and/or with weak (1+), moderate (2+), or strong (3+) membranous expression, respectively. For teratomas, all components of the tumor (i.e., epithelial and mesenchymal) were considered.
Immunofluorescence double staining was used to assess the co-expression of OCT4 and CLDN6 in selected cases to confirm diagnosis of GCNIS. 18 Slides were stained with rabbit monoclonal antibodies against human OCT4 (clone EPR17929; Abcam Limited, Cambridge, UK; 1:100) and CLDN6 (clone E7U20; Cell Signaling Technology, Danvers, MA, USA; 1:75) on a Ventana Discovery Ultra GG instrument using Opal570 and Opal690 fluorophores and counterstained with 4,6-diamidino-2-phenylindole (DAPI). The stained slides were digitized using an Akoya Phenoimager (Akoya Biosciences, Marlborough, MA, USA).
Presence of i12p was assessed in all specimens utilizing a fluorescence in situ hybridization (FISH) assay at PathWest Laboratory Medicine WA, Australia. Three micrometer unstained tissue sections were deparaffinized in three changes of Hurstol then washed in absolute ethanol. Antigen retrieval was achieved by incubating the sections in preheated (98°C) 0.01 sodium citrate solution followed by incubation in 2.5 mg/mL pepsin solution at 37°C for 15 min, then dehydrated in a series of 70%, 90%, and 100% ethanol. The sections were then hybridized at 37°C overnight with the ZytoLight SPEC KRAS/CEN 12 Dual Color probe (Zytovision, GmbH; Bremerhaven, Germany). Post hybridization washes were performed the next day in 0.4×SSC/0.3% Igepal buffer at 72°C for 2 min followed by 2×SSC/1% Igepal at room temperature for 2 min, then incubated in a DAPI solution at room temperature for 5 min to counterstain the nuclei.
The slides were examined using an Olympus BX53 fluorescence microscope equipped with Chroma technology 83,000 filter with a single- and dual-band excitors for SpectrumOrange, Spectrum Green, and DAPI (uv 360 nm), and the Cytovision imaging software (Leica Biosystems; Wetzlar, Germany). Hundred nuclei for each sample were assessed for the number of KRAS and CEN 12 signals as well as the relative positions of each of the signals. A signal pattern with two copies each of the KRAS (green) and CEN 12 (orange) was scored as negative for i12p. Nuclei which showed a signal pattern of extra KRAS signals and a close juxtaposition of two KRAS signals with a CEN 12 signal, were scored as positive for i12p. Nuclei showing additional KRAS signals without the juxtaposition of KRAS and CEN 12 was scored as negative for i12p but positive for KRAS amplification.
The reporting of this study conforms to the STROBE guidelines for observational research 19 (see Supplemental Material).
Results
Matched tumor samples collected before and after chemotherapy containing metastatic teratoma were available from 22 patients (see Table 1 and Figure 1(a)). The median age was 29 years; 8 (36%) had relapsed stage 1 GCT, and the remainder de novo advanced disease. Almost all participants had primary mixed GCTs (20/22, 91%), with embryonal carcinoma and teratoma forming the most common non-seminomatous components (each 17/22, 77%). Two pure, primary teratomas were observed (9%). All orchidectomies occurred prior to chemotherapy. At initiation of treatment for metastatic disease, most individuals (15/22, 68%) had International Germ Cell Cancer Collaborative Group good-prognosis testicular GCT.
Clinicodemographic details of patient cohort.
CDCT, conventional-dose chemotherapy; HDCT, high-dose chemotherapy; IGCCCG, International Germ Cell Cancer Collaborative Group.

Expression of CLDN6 in primary testicular GCTs and metastatic teratoma. (a) CLDN6 expression in primary GCTs (n = 22) across all components including teratoma-components. (b) CLDN6 expression in mixed primary GCTs with teratoma-component (n = 8/22) in different tumor-components (total tumor, GCT non-teratoma vs teratoma-component). (c) CLDN6 expression in primary GCTs and matched metastatic teratoma across all components, respectively (n = 22). The expression of CLDN6 is shown as percentage of positive tumor cells, and the colors indicate the staining intensities. Cases are ordered according to expression levels in primary tumors. The table below shows the FISH data (presence of i12p, KRAS amplification, and aberrations affecting chromosome 12), the percentage of teratoma and nonteratoma GCT components present on the evaluated slide, and the histological subtypes that were present in the entire tumor are depicted. In addition, the side of origin of the investigated metastases are given in (c).
Prior to postchemotherapy teratoma resection, almost all individuals (21/22, 95%) had received one line of systemic therapy only; none had received HDCT. Teratoma was observed as the sole germ cell component in resected metastases in 86% (19/22) of participants.
Histological analyses were performed in 22 primary cases; 12/22 (55%) had only nonteratoma GCT on available slides, 8/22 (36%) presented with both nonteratoma and teratoma components, and 2/22 (9%) with pure teratoma (see Figure 1(a)). Membranous CLDN6 expression was identified in 21 primary cases (21/22, 95%); 1 individual, 030-A, with pure teratoma demonstrated no CLDN6 expression (see Figures 1(a) and 2). Detailed subtype analyses of eight primary tumors with both nonteratoma and teratoma components showed CLDN6 positivity of all nonteratoma components (8/8, 100%; see Figures 1(b) and 2), whereas only 3/8 teratoma components showed CLDN6 positivity (38%; Figure 3). Likewise, metastatic teratomas rarely expressed CLDN6 (5/22, 23%), and when expressed, this was observed at low levels only (see Figures 1(c) and 3 and Table 2).

Expression of CLDN6 in GCTs without (a, b) and with teratoma-component (c, d). H&E staining is depicted on the left side, immunohistochemical expression of CLDN6 on the right side. Arrows highlight the teratoma-component in a mixed GCT (lower row). (a) H&E-stained slide showing nonteratoma GCT. (b) Expression of CLDN6 in nonteratoma GCT. (c) H&E-stained slide with nonteratoma GCT (upper left and lower part of the image) with intermixed teratoma consisting of intestinal-type epithelium and a surrounding mesenchymal component. (d) CLDN6 expression is restricted to the nonteratoma component. Arrows highlight the epithelial component of the teratoma.

Expression of CLDN6 and KRAS amplification in teratomas. (a) H&E staining showing teratoma with cystic spaces lined by squamous (lower right) and glandular epithelia with surrounding stroma. (b) FISH analysis of a teratoma sample with KRAS amplification (green signals) and lack of i12p (CEN 12; red). (c) CLDN6 immunohistochemistry with lack of expression in mesenchymal and epithelial cells. (d) Focal epithelial expression.
CLDN6 expression in primary and metastatic testicular GCTs.
The frequency of CLDN6 protein expression in 22 matched GCT tissue samples was determined using a semiquantitative IHC assay. A board-certified pathologist evaluated each sample and determined the proportion of tumor cells staining negative (0), weakly positive (1+), medium positive (2+), or strongly positive (3+) for membranous CLDN6 expression. The mean of the proportion of CLDN6-positive tumor cells are given for samples with any CLDN6 positivity (n).
CLDN6, Claudin-6; GCT, germ cell tumors; IHC, immunohistochemistry; na, not applicable.
In CLDN6-positive primary tumor samples, the mean proportion of CLDN6-positive tumor cells was 74% across all components, with a notable disparity between nonteratoma components, which exhibited 88%, and teratoma components, which showed only 7% positive cells (see Figures 1(a) and (b) and 3, and Table 2). Similarly, CLDN6 positive tumor cells were sparse in metastatic teratomas, with a mean of 5% positive tumor cells detected in the five metastatic teratomas expressing CLDN6 (see Figures 1(c) and 3). A side-by-side comparison of matched sample pairs revealed substantially lower expression of CLDN6 in postchemotherapy teratoma compared to the respective GCT in the primary orchidectomy (see Figure 1(c) and Table 2). In all cases, including primary and metastatic tumors, irrespective of histological components, tumor cells exhibited low to medium expression (1+/2+), while tumor cells with strong intensity (3+) were rare (see Table 2).
i12p was observed in 82% (18/22) of all orchidectomy specimens, including those containing pure teratoma (2/2, 100%), and of the 4 participants without i12p, 2 (9%) demonstrated other chromosome 12 aberrations (aneuploidy or polysomy). KRAS amplification was uncommon, and observed in one orchidectomy, which did not demonstrate i12p (5%; see Figure 1(a)).
In metastatic postchemotherapy teratomas, i12p was identified in 14/22 (64%), KRAS amplification in 8/22 (36%), and other chromosome 12 aberrations in 2/22 (9%) samples (Figures 1(a) and (b) and 3). KRAS amplification was acquired in 7/8 metastatic samples (88%) and was observed in conjunction with i12p in 5/8 (63%). Concordance between i12p in primary GCTs and metastases were observed in 13/18 cases (72%).
In addition, GCNIS were identified in 16 of the matched pairs (73%). GCNIS in all cases showed membranous CLDN6 expression with intensity ranging from 1+ to 3+. i12p was observed in GCNIS (see Figure 4(A)).

CLDN6 in the testicular GCT progression model. (A) H&E (upper row), CLDN6 immunohistochemistry (middle row), and FISH test for detection of i12p (lower row) in normal tissue, GCNIS, primary GCT, and metastatic teratoma (i-k). The presence of i12p isochromosome is highlighted by close association of red and green signals. While CLDN6 is negative in normal germ cells (a, e), expression becomes evident in GCNIS (b, f) and in nonteratoma GCT (c, g). It is mostly absent in teratoma (d, h). (B) Proposed progression model of CLDN6 expression during GCT tumorigenesis.
Discussion
This study, to our knowledge, is the first to characterize CLDN6 expression and the presence of i12p in a cohort of individuals with paired primary GCT and metastatic teratoma samples, as well as GCNIS. The results of our exploratory study confirm high expression of CLDN6 among nonteratoma GCT elements and highlight early expression in GCNIS, though critically, low expression in teratoma, rendering CLDN6-directed therapies unlikely to be efficacious against primary or metastatic teratoma. In contrast, i12p was observed in a high proportion of primary orchidectomy and metastatic teratoma specimens, though concordance in paired samples was lower than expected and present in just over two-thirds in our analysis. Despite lacking therapeutic actionability (no active trials ongoing to our knowledge), its potential role in detecting viable GCT and teratoma in postchemotherapy residual masses remains important, and this finding may propel the development of circulating assays to guide clinical care and aid patient selection for treatment. The absence of this biomarker in around one-third of metastatic teratoma-containing specimens, however, does raise concern regarding the clinical impact of deferring curative, postchemotherapy resection in an i12p “negative” population.
Our study provides a distinctive opportunity to investigate CLDN6 expression and the presence of i12p during the sequential progression of testicular GCTs, enhancing previous findings that demonstrate immunohistochemical expression of CLDN6 in fetal testicular germ cells and its absence in postnatal testicular tissue from birth to adolescence. 20 Similarly, our study cohort revealed no CLDN6 expression in adjacent normal testicular tissue (see Figure 4(A)). Collectively, these findings suggest a progression model characterized by early CLDN6 upregulation in noninvasive GCNIS, which persists in corresponding nonteratoma subtypes of GCT (see Figure 4(B)). Conversely, a lack of CLDN6 expression is evident in primary teratoma and in teratoma components of primary GCTs, as well as in posttreatment teratoma derived from CLDN6-positive GCTs, while i12p is consistently observed in 72% of cases throughout the stepwise tumor progression, although not universally detected.
There is an urgent need to progress predictive biomarkers and therapeutic strategies for unresectable teratoma and address this important gap in clinical care. Surgery remains the only pathway to cure due to the inherent risks of somatic transformation of postpubertal teratoma, or development of growing teratoma syndrome. Serum tumor biomarkers markers (AFP, β-hCG, LDH) and imaging offer low sensitivity for detecting teratoma. Limited insight into its molecular landscape has hamstrung progress, leaving this small but important group of patients with unresectable teratoma without effective options. Novel biomarkers are urgently needed.
Strengths and limitations
Our study has several strengths which underpin the importance of clinical cancer registries, such as iTestis, that concurrently collect high-quality clinical data and biospecimens for translational research. Real-world data are of critical importance to the practice of oncology, particularly with rare cancers like GCT, where randomized trials are not always feasible, and collaboration between clinician researchers and industry partners is key to progressing clinically relevant advances.
We present robust clinical data on a small, yet rare population with matched primary and metastatic tissue pairs, coupled with high-quality translational analyses including detailed correlation with histological subtypes and distinct analysis on nonteratoma and teratoma components. Though our analysis could be strengthened by greater numbers, and availability of representative tumor blocks containing both nonteratoma and teratoma component in all orchidectomy samples, if applicable, the low levels of CLDN6 expression observed in the two pure teratoma orchidectomies, and the eight samples in which teratoma was scored separately, provides reasonably conclusive evidence that CLDN6 expression is low in this histological GCT subtype.
Conclusion
While GCNIS and nonteratoma GCTs exhibit both frequent and high expression of CLDN6, the minimal to absent expression in primary and posttreatment teratoma suggests that CLDN6-targeted therapies are unlikely to be clinically beneficial for teratoma management. These findings underscore the need for alternative therapeutic strategies tailored to the unique challenges posed by teratomas.
Supplemental Material
sj-docx-1-tam-10.1177_17588359261421810 – Supplemental material for Evaluating Claudin-6 and isochromosome 12p in the progression from germ cell neoplasia in situ to primary testicular germ cell tumor and post-chemotherapy teratoma: implications for CLDN6-targeted therapies
Supplemental material, sj-docx-1-tam-10.1177_17588359261421810 for Evaluating Claudin-6 and isochromosome 12p in the progression from germ cell neoplasia in situ to primary testicular germ cell tumor and post-chemotherapy teratoma: implications for CLDN6-targeted therapies by Ciara Conduit, Bruno Valentin Sinn, Catherine Mitchell, Jonas Leichsenring, Mark Laible, Kristina Zlatic, Miku Kuba, Sophie O’Haire, Stefan Wöll, Anna Melissa Schlitter and Ben Tran in Therapeutic Advances in Medical Oncology
Footnotes
Acknowledgements
Florian Frohns, Astrid Houben, Jennifer Coridaß, and Samira Rudlaff (BioNTech SE) assisted with the planning and conducting of immunohistochemical analyses. We thank all the patients who have donated their tumor/tissue for research.
Declarations
ORCID iDs
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
