Abstract
Background/Aim: Adult granulosa cell tumor (aGCT) is a rare and challenging ovarian tumor due to its unpredictable recurrence and its associated increased risk of breast and endometrial cancer. Identifying and describing molecular alterations in tumors has become common with the advent of high-throughput sequencing. However, DNA sequencing in rare tumors, such as aGCT, often lacks statistical power due to the limited number of cases in each study, thereby clinical implications of DNA alterations are difficult to interpretate. This scoping review aims to systematically describe somatic and germline DNA alterations identified in women with aGCT. Materials and Methods: Search terms (granulosa cell tumour AND molecular alterations) were searched in May 2024 in the following databases: MEDLINE (Ovid), Embase (Ovid), Web of Science Core Collection and Google Scholar. Screening, full-text review and data extraction were performed by two independent reviewers. Results: Twenty-four publications were identified. Eighteen reported on somatic DNA alterations of patholgenic mutations identified in total 1,226 tissues being sequenced. FOXL2 (c.402C>G; p.C134W) was present in 97% of aGCTs. Other pathogenic mutations in the tissues investigated were TERT promoter mutation (41%), truncating KMT2D mutations (14%) and TP53 pathogenic variant (4%). TERT promoter mutation was reported more frequently in recurrent tumors (p<0.01), whereas comparing truncating KMT2D and TP53 mutations reported in primary and recurrent tumors revealed no difference (p=0.15 and p=0.26 respectively). Tumor mutational burden (TMB) was reported in five studies and all showed a low TMB. None of the somatic mutations were candidate targets for biological treatment. Six publications reported germline variants and no shared germline pathogenic variants were described in the published literature. Conclusion: The FOXL2 missense mutation was the only common somatic DNA alteration in aGCT. TERT promoter mutations were reported more frequently in recurrent aGCT but their clinical relevance remains uncertain. In contrast to previous reports, truncating KMT2D mutations were not found to be associated with recurrent aGCT. Evidence on common germline variants in aGCT is sparse. The role of somatic and germline DNA alterations in the development of other malignancies in women with aGCT remains uncertain. Further research involving matched primary and recurrent tumors, as well as other primary malignancies, is essential to better understand the mutations that drive tumor development.
Ovarian adult granulosa cell tumor (aGCT) is a rare neoplasm originating from the ovarian stroma, accounting for approximately 2-5% of ovarian malignancies (1, 2). Most aGCTs are diagnosed at an early stage and can be curatively treated with surgical resection. About 10-20% of cases relapse with advanced tumor spread, sometimes many years after the initial diagnosis (3, 4). Additionally, aGCT patients have an increased lifetime risk of developing other cancers, particularly estrogen-sensitive cancers, such as breast and endometrial cancer (5, 6). While the molecular landscapes of breast and endometrial cancers have been extensively studied, no common germline or somatic alteration have been identified across breast, endometrial, and aGCT (7).
Managing aGCT presents certain challenges. Beyond the tumor stage, no prognostic biomarkers are currently used to predict potential recurrences (8, 9). Evidence-based treatment options for recurrent aGCT are lacking, with further surgery being the primary approach, and no significant results with targeted therapies have been reported (10-15).
A missense mutation in FOXL2 (c.402C>G; p.C134W) has been reported in approximately 95% of aGCT cases (16, 17). However, while this FOXL2 mutation is valuable for diagnosis, its clinical relevance remains limited (18).
Due to the relative rarity of the tumor type, molecular investigation of aGCT is so far limited to studies with small case numbers and are not sufficient to describe the linkage between the genomic landscape and prognostic or actionable targets. Although several reviews on aGCT exist, no one has systematically mapped the current knowledge of somatic and germline DNA alterations in aGCT(19-26).
This scoping review aims to systematically describe somatic and germline DNA alterations found in women with aGCT and determine the clinical implications of these alterations regarding recurrencies, targeted treatment eligibility, and associations with the development of other primary malignancies.
Materials and Methods
The methods used have been described previously (27). This review was conducted in accordance with the methodology outlined by Arksey and O’Malley (28), and the Joanna Briggs Institute (29). We adhered to the PRISMA Extension for Scoping reviews (PRISMA-ScR) checklist (30).
Search strategy. The search strategy consisted of two search terms (granulosa cell tumour AND molecular alterations). Each element consisted of database-specific subject headings, and free text words with relevant truncation and proximity operators. The full search syntax is available in the published protocol (27). MEDLINE, Embase (Ovid), Web of Science and Google Scholar (100-top ranked) were searched. The search was performed in May 2024.
Study selection. Inclusion criteria was: i) peer-reviewed original research focusing on human aGCT and somatic/germline DNA mutations/alterations, ii) articles defining aGCT-diagnosis as validated by pathologists prior to molecular analysis, and iii) articles describing sequencing results of somatic and/or germline DNA alterations in women diagnosed with aGCT. There were no publication date restrictions, and all studies matching our criteria published up until the search date were included. Exclusion criteria was: i) published results limited to FOXL2 (c.402C>G; p.C134W) known to be present in ~95% of aGCT, irrespective of technology platform used, were excluded and ii) published results derived from cell lines only. Results were exported to Covidence (31) for screening, and any duplicates were removed. Two reviewers (SK and FFL) screened and reviewed the publications. Any disagreements were resolved by discussion, or if necessary, an experienced third author (TSP, EH) made the final decision. Study selection is outlined in the PRISMA-flowchart (Figure 1).
Preferred reporting items for systematic reviews and meta-analysis (PRISMA) study flow diagram for selection. aGCT, Adult granulosa cell tumor.
Data extraction and quality assessment. The risk of bias was not evaluated because the included reports were purely descriptive. The following patient characteristics were extracted from the included studies regarding somatic DNA mutations: histology, tumor mutational burden (TMB), and DNA alterations, if reported. Any somatic DNA mutations, known as either pathogenic or likely pathogenic according to OncoKB Cancer Gene List were recorded (32). Histology of tissue used in the articles describing somatic mutations was classified as either primary, recurrent tissue or not reported status of aGCT tissue used for analyses. The proportion of somatic DNA mutations was calculated based on the total number of samples in articles that were specifically sequenced. Somatic alterations were reported in this review if they were altered in 1% or more of all patients and investigated in at least 10% of the aGCT cases.
The following patient characteristics were extracted from the included studies regarding germline DNA variants: type of sample, sequencing method, pathologic variants, if reported. Due to small number of publications, all published results on germline variants in women with aGCT were recorded.
For both somatic mutations and germline variants, raw sequencing results were used as basis for data extraction, if provided by the publisher as supplemental material. Otherwise, data were extracted directly from the Materials and Methods or from the Results sections of the articles. If the original article did not report the data, it was recorded as not reported.
To determine if any of the reported somatic or germline DNA mutations/variants are associated with other primary malignancies and if these mutations/variants are eligible for targeted treatment, public databases (MyCancerGenome.org, OnkoKB, ClinVar, COSMIC, ClinicalTrial.gov) were referenced.
Categorical data were compared using χ2-test on STATA Release 17.0 (StataCorp, College Station, TX, USA).
Results
Our search retrieved a total of 1,795 records of which 111 were included in full text review based on title and abstract screening (Figure 1). Twenty-four publications were selected based on the inclusion and exclusion criteria. Nineteen publications reported on somatic DNA mutations of which nine included analysis of matched blood or normal tissue. Five publications reported solely on germline variants in women with aGCT (Table I and Table II).
Overview of the included articles regarding somatic DNA mutations in ovarian adult granulosa cell tumor.
Overview of published results of germline DNA variants in women with ovarian adult granulosa cell tumor.
The sequencing techniques employed included Sanger sequencing, TaqMan Single Nucleotide Polymorphism Genotyping used for detection of TERT promoter mutations, targeted next-generation sequencing (NGS), whole exome sequencing (WES), and whole genome sequencing (WGS). The 24 included studies comprised 1226 tumor samples sequenced for somatic DNA mutations and 266 samples sequenced for germline DNA variants.
Somatic DNA mutations. A total of 1,226 tumor samples were analyzed in the included publications. Of these 462 (38%) were classified as primary tumors, 206 (17%) as tissue from recurrent disease/metastasis and 558 (45%) were unclassified regarding the origin of primary or recurrent status (none reported status). Figure 2 displays the eight identified somatic DNA alterations which occurred with a frequency of >1%. FOXL2 (c.402C>G; p.C134W) mutation was present in 97% of all sequenced samples. TERT promoter mutation was sequenced in 993 samples and a pathogenic mutation was identified in 41% of the investigated samples. Truncating mutation in KMT2D was present in 14% of the 704 sequenced samples.
Somatic DNA mutations present in >1% of ovarian adult granulosa cell tumor samples found in the present review. A total of 1,226 tumor tissue samples were identified; however, due to the different sequencing methods used, the number of samples analyzed for each somatic mutation varies and does not sum to the total. *Classified as either pathogenic or likely pathogenic.
TERT promoter mutation was identified more often in recurrent tumors than in primary (46% vs. 27%, p<0.01), whereas there was no difference when comparing primary to recurrent aGCT regarding truncating KMT2D mutation (8% vs. 13%, p=0.15) or TP53 mutation (3% vs. 6%, p=0.26) (Figure 3). FOXL2 (c.402C>G; p.C134W) was similarly frequent in recurrent tumors as compared to primary (98% vs. 97%, p=0.2).
Comparison of DNA mutations found in primary and recurrent ovarian adult granulosa cell tumors. After FOXL2 c.402C>G, TERT promoter, truncating KMT2D, and TP53 mutations were the most frequent somatic mutations. P indicates primary aGCT tumor tissue and R indicates recurrent aGCT tissue. These samples are not matched. *p-Value was calculated by χ2-test.
Three publications aimed to analyze both primary and matched recurrent aGCT [primary, n=37; recurrent, n=35 (2 missing in Alexiadis et al.)]. For all patients, somatic FOXL2 mutation was present in both primary and matched recurrent tumor tissue. TERT promoter mutation was present in tissue from 14 primary and 22 recurrent samples, respectively (p=0.04) (33-35).
TMB was reported in only five studies and the results were reported as low TMB (0.46-3.0 mutations/Mb) for all investigated samples.
Germline DNA alterations. A total of 266 blood samples were recorded in 14 studies and most samples (n=252) were sequenced to validate the results of somatic DNA mutations in aGCT (33, 34, 36-41). Six studies reported on pathogenic germline variants (Tabel II). None of the reported pathogenic variants were shared or common amongst women with aGCT. In the published literature, two women with aGCT were carriers of a germline pathogenic variant in the BRCA1 or BRCA2 genes. The eight studies which included sequencing of blood as a reference for somatic sequencing results reported no pathogenic germline variants.
Actionable targets. None of the described somatic mutations have approved targeted therapies in the standard treatment of aGCT according to the European Medcines Agency (42). For alterations in the genes CDKN2A, PIK3CA, and MTAP, approved targeted therapies are available for cancers of another origin (32, 42). Mutations in these genes were described in 27/573 (5%), 26/673 (4%) and 12/572 (2%) of sequenced aGCT tissue samples, respectively. One trial (ALEPRO, NCT05872204), which examines the effectiveness of abemaciclib (CDKN2A and PIK3CA) and letrozole in patients with estrogen receptor-positive rare ovarian cancers (43), is still ongoing, including estrogen positive low grade serous ovarian carcinomas as well as aGCT. It was initiated in November 2023 and inclusion of patients is estimated to end in November 2026. No results are published yet.
Discussion
This scoping review found that FOXL2 (c.402C>G; p.C134W) was present in almost all of the sequenced aGCT tissue samples (97%) in the published literature. TERT promoter mutation was seen more frequently in recurrent tumors (primary 27%; recurrent 46%, p<0.01). Truncating KMT2D mutations, previously described as associated with recurrent tumors, were not found to be more common in recurrent tumors in this systematic mapping of DNA alterations in aGCT (p=0.15). No significant number of germline DNA variants or somatic DNA mutations were found in aGCT linked to an increased risk of other primary malignancies. No targeted therapy for aGCT is currently available as standard treatment, and only one clinical trial is ongoing. Somatic mutations in the genes CDKN2A, PIK3CA, and MTAP were described in 5%, 4%, and 2% of aGCT tissue samples, respectively. The mutations reported in these genes are eligible for targeted treatment in other malignancies; however, experience with such treatments in aGCT is lacking.
A somatic missense mutation in FOXL2 (c.402C>G; p.C134W) is found in most aGCTs, and in this review we further confirm this. The exact mechanisms of FOXL2-mutation in aGCT disease development still remain uncertain. Shin et al. (44) have investigated whether FOXL2 (c.402C>G; p.C134W) acts as an oncogene or a tumor suppressor gene and suggested that miR-1236 selectively degrades the FOXL2 c.402C>G mRNA, leading to somatic haploinsufficiency, implying its potential role as a tumor suppressor (44). Pilsworth et al. argued against FOXL2 c.402C>G behaving as a tumour suppressor, as this would depend on an allelic imbalance that was found neither in the samples included in their study nor in the samples analyzed by Shah et al. (16, 45). Three recent studies have examined the oncogenic mechanisms associated with FOXL2 c.402C>G mutation (18, 46, 47). One of them suggested that the alteration drives aGCT by altering the binding affinity of FOXL2-containing complexes to engage an oncogenic transcriptional program (46). Another study found that FOXL2 c.402C>G hijacks SMAD4, promoting the expression of genes linked to epithelial-to-mesenchymal transition, stemness, and oncogenesis (18). Consistently, a recent study showed that the FOXL2 allele underlying C123W was dominant regarding spontaneous aGCT development through gain of function in a mouse model (47). Combined, these papers suggest that FOXL2 (c.402C>G; p.C134W) is an oncogene.
Understanding the FOXL2 c.402C>G pathway in disease development is crucial for developing targeted treatment options. However, no reported benefits of targeted treatment in FOXL2 mutated aGCT in humans has been published. An in vitro experiment on an immortalized aGCT cell line (SVOG3e), showed sensitivity of the cells to YM155 (survivin suppressant), driven by FOXL2 c.402C>G (46). Weis-Banke et al. suggested a potential benefit of TGFβ or activin-inhibitors in aGCT based on their findings that TGFβ-inhibition dramatically antagonizing the transcriptional effect of the FOXL2 C134W/SMAD complex (18). Further studies are needed to elucidate the potential of targeted pathway inhibition in FOXL2 c.402C>G.
FOXL2 c.402C>G might also be useful in aGCT disease monitoring, as supported by Groeneweg et al. They demonstrated that the levels of FOXL2 mutant circulating tumor DNA found in the plasma of the patients can predict disease progression, as well as treatment response (48). Larger studies are warranted to validate these results and to investigate whether earlier detection of progression impacts overall survival.
The absence of FOXL2 c.402C>G in a small subset of aGCTs suggests the existence of FOXL2 wild-type aGCTs (16). A multicenter study retrospectively validated aGCTs (n=336) that had previously undergone FOXL2 c.402C>G mutation analysis (49). The study found a misdiagnosis rate of approximately 18% in FOXL2 wild-type aGCT cases, of which some were reclassified as epithelial ovarian malignancies. This underscores the importance of performing mutational analysis of FOXL2 in aGCT, as cases with FOXL2 wild-type should receive additional scrutiny during pathological review.
TERT promoter mutation is found in more than 50 different cancer types, and is a common mutation in many advanced and metastasized cancers (50, 51). In this review, we found somatic TERT mutations (C228T or C250T) to be present in 405/993 aGCT tissue samples. A TERT promoter mutation was found more frequently in recurrent aGCT tissue and may therefore be an important event to facilitate aGCT-recurrence (34, 35). However, studies in which TERT promoter mutation is determined in both primary and subsequent recurrent aGCT are lacking. In a Dutch study, TERT promoter mutant circulating tumor DNA was monitored in three women and found to be correlated with disease progression (48). Larger studies are warranted to validate these results and to investigate the impact of TERT promoter mutation in a clinical context.
Truncating KMT2D mutation is frequent in cancers, most commonly associated with follicular lymphoma and diffuse large B-cell lymphoma (52). Hillman et al. proposed that truncating KMT2D mutation in aGCT is found more frequently in recurrent aGCT, suggesting that this mutation is a driver in disease progression (36). However, this study lacked matched primary and recurrent tumors, making it difficult to determine if truncating KMT2D mutations were already present in the primary tumors. In this review, we found no difference in the frequency of truncating KMT2D mutations in recurrent compared to primary aGCT. Studies with matched primary and recurrent tumors are needed to further elucidate the role of truncating KMT2D mutation in aGCT.
As described in a previous scoping review with the aim to map molecular alterations in rare cancers, a scoping review of this sort has some limitations (53). One limitation is due to the heterogeneity of the included studies and the limited number of samples investigated in each study. Another limitation of the included studies in this review is that the tumor samples were drawn from select cohorts at tertiary academic centers, likely representing a patient population with more advanced disease. It is crucial to note that the studies reviewed here involved diverse populations of aGCT patients, including treatment-naive patients and those treated with adjuvant chemotherapy. Additionally, caution should be exercised when comparing DNA-sequencing results from different studies that employ varying sequencing methods. Targeted NGS studies, which explore a specific set of genes may not directly be comparable to WGS studies that cover the entire genome. Processing data from high-throughput sequencing requires bioinformatics pipeline frameworks, which are essential for identifying clinically relevant genetic alterations. These frameworks vary between studies, leading to differences in sensitivity for detecting genomic alterations (54). The different sequencing technologies furthermore have different sensitivity, which should also be considered in the comparison of the reported results. The same variability applies to the measurement of TMB being a surrogate biomarker of genomic instability, where bioinformatics algorithms significantly influence the results (53, 55).
Another major challenge in interpreting impact of DNA-sequencing in aGCT samples is the limited number of studies that have analyzed both primary and matched recurrent aGCT tissues alongside primary non-recurrent tumors. We identified only one study that examined a subsample of primary non-recurrent aGCT alongside primary aGCT with matched recurrent samples (34).
In summary, this scoping review highlights that aGCT is predominantly a FOXL2-centric disease for the vast majority of patients, with only a few DNA mutations reported as significant for recurrent disease. The molecular mechanisms of the FOXL2 c.402C>G mutation in disease progression need further elucidation. Currently, no targeted therapy for aGCT is available as a standard treatment. No somatic DNA mutation or germline DNA alterations in aGCT appear to contribute to the development of other primary malignancies. For cases of concurrent aGCT and other primary cancers, such as endometrial or breast cancer, future research should analyze matched cancer tissues to determine whether these cancers share molecular characteristics that contribute to cancer development. Additionally, larger cohorts with matched primary and recurrent aGCTs are necessary to clarify the roles of TERT promoter mutations and truncating KMT2D mutations in disease progression.
Footnotes
Authors’ Contributions
Sven Karstensen: Conceptualization, Methodology, Validation, Formal analysis, Writing - Original Draft. Karsten Kaiser: Validation, Writing - Review & Editing. Tim Svenstrup Poulsen: Validation, Writing - Review & Editing. Kirsten Jochumsen: Writing - Review & Editing. Claus Høgdall: Writing - Review & Editing. Niels Marcussen: Writing - Review & Editing. Finn Friis Lauszus: Validation, Writing - Review & Editing, Supervision. Estrid Høgdall: Conceptualization, Methodology, Validation, Writing - Review & Editing, Supervision.
Funding
The present study was funded by the Region of Southern Denmark Health Administrations Ph.D. scholarship.
Conflicts of Interest
The Authors declare no competing interests.
- Received October 22, 2024.
- Revision received November 18, 2024.
- Accepted November 19, 2024.
- Copyright © 2025 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.