Molecular and histopathological characterization of seminoma patients with highly elevated human chorionic gonadotropin levels in the serum

Germ cell tumors (GCTs) comprise over 95% of testicular cancers and account for approximately 75.000 new cases and over 9.000 deaths worldwide each year [1]. Despite being a rare tumor entity, testicular cancer is the most common malignancy in male adolescents and young adults and its incidence has been increasing during recent years [2, 3]. As GCTs originate from an early developmental stage, they divide into seminomas (SEM) and non-seminomatous germ cell tumors (NSGCT). Characteristically, GCTs, especially non-seminomas (non-SEMs), harbor a broad phenotypic heterogeneity due to a differentiation potential based on pluripotency features of the stem cell-like population of embryonal carcinoma [4]. Both types originate from testicular germ cell neoplasia in situ (GCNIS), which is considered to originate from developmentally arrested primordial germ cells (PGC) [4]. SEMs resemble masses of GCNIS, appearing morphologically close to PGC and proliferate as homogeneous tumors, retaining features of germinal lineage. Approximately 30% of SEM patients with advanced disease present with moderately elevated serum levels of beta human chorionic gonadotropin (hCG). The possible presence of this diagnostic marker in SEM is well documented and accounted for the presence of syncytiotrophoblastic cells (STC) [5, 6]. However, in these cases, serum hCG levels rarely exceed 500 IU/L and extraordinarily high hCG levels have only been described in few cases [7, 8]. Therefore, in case of high hCG levels the presence of a non-SEM component in the tumor, i.e. choriocarcinoma (CC), may be suspected. To better characterize hCG⁺ SEM patients, we previously conducted an international registry with the German Testicular Cancer Study Group in collaboration with the Global Germ Cell Tumor Collaborative Group (G3) analyzing clinical features from 1031 patients with SEM and hCG levels above normal at first diagnosis according to local laboratory ranges [9]. A small subset of patients in this registry presented with extraordinarily high hCG levels at initial diagnosis, suggesting that this subgroup may represent a distinct prognostic category, as these patients had an unfavorable outcome. Recently, Wagner and colleagues described as an independent predictor for relapse [10]. Since this retrospective registry provided no central pathological review, the question was raised whether non-SEM components were present in the tumor specimens of these patients, therefore suggesting an incorrect diagnosis of classical SEM when NSGCT was present. A correct diagnosis and stratification of GCTs in SEM and NSGCT is important, because treatment and prognosis can differ for these entities [11]. Consequently, we conducted a comprehensive analysis of tissue and RNA material from patients from this registry diagnosed with SEM with excessively elevated serum hCG levels defined as > 1000 U/L. The aim of this study was to validate the diagnosis of pure SEM in patients and to find a rationale for the highly elevated hCG levels in these patients.

Excessively elevated serum hCG at the time of diagnosis in SEM is an uncommon finding and therefore clinicians often suspect misdiagnosis of pure SEM in the presence of NSGCT components. As a testicular tumor with SEM and present NSGCT components is classified as NSGCT, this can imply prognostic and therapeutic consequences for the affected patients. In this study, we analyzed pure SEM with very elevated serum hCG levels over 1000 U/L, to confirm the diagnosis and find an explanation for high serum hCG levels in these cases. Elevated serum hCG in SEM with the presence of syncytiotrophoblasic giant cells is a known phenomenon, however, these cases usually appear with moderate serum hCG levels not exceeding 500 U/L. In this study, we analyzed tissue and RNA samples of SEM specimens in patients with serum hCG levels of up to 40,909 U/L. Central histomorphological analysis revealed typical findings of SEM in all patients and comprehensive IHC analysis proved a classical SEM-like phenotype with the presence of syncytiotrophoblastic giant cells. Moreover, an analysis of RNA of SEM from patients with high hCG levels revealed a typical gene expression signature of SEM with the presence of syncytiotrophoblastic giant cells. Importantly, since we could exclude presence of EC next to the trophoblast cells or within the SEM tissues in general, we hypothesize that SEM are able to undergo extra-embryonic differentiation into trophoblast/CC lineage, a concept already stated by Ulbright et al., but now strongly supported by our data [20].

Interestingly, our analysis of the TCGA data of mixed GCT with high hCG expression revealed that most of the analyzed trophoblast/CC marker genes were expressed, suggesting that either the CC component in hCG⁺ SEM tissues is too small to allow for detection of related markers in a bulk tumor tissue analysis or the trophoblastic / CC component in SEM is different than in NSGCT—a research question that might be addressed in the future.

An important limitation of this study is the relatively small cohort size of histologically and molecularly investigated SEM cases with very high serum hCG levels. The conclusion of the study, that all cases with exceptionally high hCG levels are secondary to the presence of syncitiotrophoblasts in seminoma is supported by a limited number of cases. Further validation is therefore needed. However, to our knowledge, this is the first study systematically reporting findings in tissue and RNA samples in this rare entity.

Taken together, our findings show that excessively increased hCG serum levels can appear in patients with pure SEM. To explain high hCG serum levels, it is important to diagnose the subtype of a SEM with syncytiotrophoblast giant cells. Due to the absence of EC, YST and TER, but presence of SEM and trophoblast/CC markers, we postulate that these hCG⁺ cells resemble trophoblast-like cells, which arose from extra-embryonic differentiation of SEM cells.