Abstract
Fertility-sparing treatment (FST) for endometrial carcinoma (EC) is an option for a subgroup of young women with low-risk disease. The low-risk group comprises patients with endometrioid EC stage IA, grade 1, with or without focal lymphovascular invasion. In the era of molecular subtyping, treatment de-escalation for some EC subtypes is recommended. Recommendations for fertility-preserving treatments were developed regardless of the molecular classification of EC. However, few studies have focused on this topic. In this review, we summarize the actual data available in the literature and discuss the impact of some molecular subtypes of FST.
Endometrial carcinoma (EC) is a serious health concern worldwide. It is one of the most frequently diagnosed malignancies among women. In Europe, according to the World Health Organization (WHO), the incidence of EC in 2020 was estimated to be 4.5% and the mortality represented 3.4% of all malignancies in women. While the majority of cases affect postmenopausal women, EC is not uncommon in premenopausal women (1, 2). Approximately 15% of these tumors arise in women aged <50 years old (Figure 1) (1).
Estimated incidence of endometrial cancer in the world according to the World Health Organization in 2020 in all ages (A) and in women aged between 20 and 49 years old (B).
The standard treatment for localized EC is surgery, which consists of total hysterectomy with bilateral salpingo-oophorectomy and lymph node dissection. Furthermore, adjuvant treatment has been tailored to histopathological tumor characteristics. These standards have major impacts on the quality of life of these patients and represent new challenges for gynecological oncologists. Recent guidelines emerging from the European Society of Gynecological Oncology (ESGO), the European Society for Radiotherapy & Oncology (ESTRO), and the European Society of Pathology (ESP) have pointed to treatment de-escalation for low-risk groups. These groups were defined based on a definitive pathological report, and comprise endometrioid EC stage IA, grade 1, with or without focal lymphovascular invasion (LVI), stage I or II POLE mutated EC, and stage IA mismatch repair deficiency (MMRd)/non-specific molecular profile (NSMP) endometroid EC, grade 1 with or without focal LVI (2-5).
The ESGO ESTRO ESP 2021 has also highlighted the importance of work-up for fertility preservation treatments and the management and follow-up of fertility preservation in young patients. Fertility-sparing treatment (FST) in EC could be an option for a subgroup of women selected based on a thorough evaluation of their reproductive potential (6). This strategy should be applied only to women with early stage non-metastatic disease. Therefore, scrupulous patient evaluation is required before surgery. A multidisciplinary team of gynecological oncologists, fertility specialists, pathologists, and radiologists must discuss each fertility-sparing strategy. Magnetic resonance imaging or transvaginal ultrasound performed by a specialized radiologist must be used to assess myometrial invasion. A second opinion from an experienced gynecopathologist is recommended. Histological assessment and grading are the gold standards for the evaluation of these patients. It has also been recommended that the G1, G2, and G3 grading system is better than the two-tiered WHO proposed system (high/low grade) (6-8).
In summary, fertility preservation options concern low-risk localized EC with a good prognosis. These tumors represent the majority of cases at the time of diagnosis of young patients. The actual fertility-sparing strategies have considered only the group of patients with endometrioid EC stage IA, grade 1, without myometrial invasion, and without other risk factors. Because there is limited evidence for grade 2 endometrioid EC, FST should be discussed on a multidisciplinary board (9). This recommendation has been developed regardless of emerging prognostic data from the molecular classification of EC.
In 2013, the Cancer Genome Atlas (TCGA) research group published data from genomic and proteomic analyses of EC. This analysis made it possible to establish four molecular subtypes associated with progression-free survival and overall survival, comprising an ultramutated subtype characterized by POLE gene mutations and a very good prognosis, a hypermutated subtype characterized by MMRd and an intermediate prognosis, a copy-number low subtype with an intermediate prognosis, and a high copy-number subtype, often associated with TP53 mutations and a worse prognosis (10).
The Proactive Molecular Risk Classifier for Endometrial Cancer (ProMisE) retrospective cohort also identified the four TCGA-based molecular subtypes for EC using immunohistochemistry and sequencing of the POLE exonuclease domain using formalin-fixed paraffin-embedded tumor blocks. The respective four sub-groups included those with MMRd, POLE mutations, and wild-type or aberrant immunohistochemical p53 expression (p53wt or p53abn, respectively). These subtypes correlate with prognosis in terms of progression-free survival, overall survival, and cancer-specific survival. Thus, favorable outcomes for POLE mutations, intermediate outcomes for NSMP and MMRd, and unfavorable outcomes for TP53 mutations have been demonstrated (Figure 2) (11, 12). Nevertheless, the ProMisE classification did not consider the tumor grade, thus limiting its use in FST, which is based only on pathological grade and tumor stage.
Algorithm for the molecular subtyping of endometrial carcinomas and their corresponding prognosis. Molecular testing for POLE mutation can be performed using polymerase chain reaction or next generation sequencing. MMR testing can be performed using immunohistochemistry or polymerase chain reaction. Abnormalities in the expression of p53 are investigated using immunohistochemistry. POLE: Polymerase epsilon; MMRd: mismatch repair deficient; NLMP: non-specific molecular profile.
Although the methods used for this classification are rapid, reproducible, and less expensive, few studies have focused on defining the molecular landscape of EC in young women. In the English literature, four studies evaluated ProMisE subtype distribution among women with EC aged <50 years. In this group of patients, the most prevalent subtype was wild-type p53, followed by the MMRd group, and finally the POLE mutated and p53abn groups (13-17). According to TCGA, the most frequently mutated genes in young women with EC were PTEN, PIK3CA, CTNNB1, TTN, and ARID1A (Figure 3). The impact of these molecular alterations on the management of EC in young women is yet to be defined. Nevertheless, some of these alterations, such as ARID1A, TSC2, CTNNB1, and HER2, could be interestingly targeted (17).
ProMisE subtype distribution and distribution of the most frequently mutated genes in women with endometrial cancer aged less than 50 years old. A) ProMisE subtype distribution across women aged less than 50 years old adapted from Ran et al. (16). B) Distribution of the most frequently mutated genes in endometrial carcinomas in women <50 years old in The Cancer Genome Atlas database. POLE: Polymerase epsilon; MMRd: mismatch repair deficient.
It is now recommended that molecular classification should be considered for all ECs, and should be performed routinely in all high-grade tumors (18). It is also recommended that molecular classification be incorporated into standard pathology reporting and treatment decision-making algorithms. These recommendations should profoundly affect fertility conservation (3-5, 19).
The aim of this review was to discuss the impact of the molecular classification of EC on FST strategies in stage I EC and to highlight its new challenges.
p53abn Endometrial Cancer Is Inconsistent With Fertility Sparing Strategies
p53abn EC occurs in older women and represents the most aggressive and lethal molecular subtype, responsible for 50-70% of EC mortality although it accounts for 15% of all EC cases (19, 20). This subtype is rarer in young women, representing 0 to 6.7% of all ECs (16). Characterized by TCGA as having a very high number of somatic copy number alterations, low mutation rate, and ubiquitous TP53 mutations (10), p53abn EC is now identified by mutant-pattern p53 immunohistochemical staining, which has been shown to be an excellent surrogate marker for TP53 mutational status (20-22). Most serous carcinomas are p53abn; however, p53abn EC may be observed across all EC histological types (21).
For a small group of tumors referred to as “multiple classifiers”, harboring more than one molecular classifying feature, specifically those with an MMRd-p53abn or POLEmut-p53abn profile, there is evidence in support of categorizing them as single classifier MMRd or POLEmut, since their outcomes correspond to those predicted by the driver molecular subtype. These tumors may be considered POLE or MMRd and treated accordingly (19, 23).
The majority of women are diagnosed after the age of 40 (90.9%) and it has been established that these tumors tend to progress in older women. A large number of these tumors were ECs (72.7%), evenly distributed between different histological grades. Myometrial invasion is observed in 60% of the cases and the presence of LVI in 40%. The hazard ratio is 12.8 for overall survival and 24.2 for disease-specific survival (using p53wt group as a reference) (13).
In a few studies that specifically addressed FST and molecular classification, the number of p53abn patients was extremely low (3/106). A complete response was observed in two of these patients and the last patient had recurrence and had undergone hysterectomy. Hence, it is currently not possible to draw definitive conclusions; however, it seems reasonable to exclude such patients (9, 15, 16, 24).
NSMP Endometrial Cancer: A Heterogenous Group to Be Refined
NSMP ECs are characterized by a low number of somatic copy number alterations, a low mutational burden, and high levels of estrogen and progesterone receptor expression. As implied by their name “no specific molecular profile”, they are defined by their lack of pathogenic POLE mutations, mismatch repair, or p53 abnormalities. They are the most common molecular subtype, accounting for approximately 50% of all ECs, with an intermediate prognosis (10, 12).
In younger women, they also represent the most frequent subtype accounting for 46.7% to 78.9% of EC (13-16). 35% of the cases are diagnosed before the age of 40 years. This subtype comprises a large proportion of ECs (97.4%) with the presence of LVI in 5% and nodal invasion in 4% of the cases. Most of the patients are stage FIGO I (85%) and classified as grade 1/2 according to the WHO classification. Their prognosis was widely debated in the literature; however, in a recent study it was demonstrated that these tumors seem to have an identical prognostic risk as the MMRd subtype, with a recurrence rate of 29.9% (16). Among the patients who underwent FST, 77 were classified as having NSMP EC. The recurrence rate is estimated to be 30% and hysterectomy is performed in 42% of cases (13). Dagher et al. reported recently a study of 20 young patients who underwent FST. In their cohort, the NSMP tumors had the best response to progesterone treatment, with 63% having a complete response (25).
Despite these new insights, this molecular subtype remains the most challenging, given the lack of predictive biomarkers that can identify patients who may benefit from personalized therapies. Recent studies have shown that this subtype could be further refined using new biomarkers, such as CTNNB1 mutations, L1CAM over-expression, and characterization of the immune infiltrate (26, 27). In this context, loss of PTEN expression has been proven not to be useful as a predictive marker of response to conservative therapy (28). These biological markers are not yet integrated to the ESGO-ESTRO-ESP recommendations and do not have any impact on treatment de-escalation (3-5, 9).
Fertility sparing in this subtype of tumor should be discussed considering the histological type, grade, and stage of the tumor. Further studies are required to better identify these patients.
Mismatch Repair Deficient Endometrial Carcinoma: Two Groups of Patients?
The MMRd/microsatellite unstable molecular subtype accounts for 25-30% of all ECs and is associated with an intermediate prognosis (10, 12). MMRd can be identified through microsatellite instability testing or immunohistochemical testing for the loss of expression of one or more mismatch repair proteins (MLH1, PMS2, MSH6, PMS2) (29). Approximately 90% of MLH1/PMS2 loss is due to somatic MLH1 promoter hypermethylation. Germline mutations in one of the mismatch repair genes, termed Lynch syndrome, account for approximately 10% of MMRd EC and 3% of all ECs. Immunohistochemistry has become mandatory in identifying patients at high risk for Lynch syndrome, so that they can benefit from appropriate counselling regarding the risk of developing additional cancers (30, 31).
Recently, Lynch Syndrome within MMRd EC was found to have an improved recurrence-free survival compared with MLH1 hypermethylation; however, this finding has no impact on treatment strategies (31, 32).
In young women, the MMRd was diagnosed in 85% of the cases after the age of 40 years. The most frequent histological subtype was endometrioid carcinoma (86%) with a predominance of the WHO G1/2 grades. Myometrial invasion was common (86%), but most of the cases were FIGO stage I. LVI was present in 60% and nodal metastasis was found in 20% of the cases (13).
After surgery, the risk of recurrence is higher in patients with MMRd status than in those with wild-type p53 tumors (13). In young patients, there were no significant differences in terms of complete response and recurrence rates between the MMRd and p53wt subtypes (16). However, recent studies by Chung et al. and Dagher et al. found that patients with MMRd had poor outcomes after hormone therapy for early stage EC. Although these results were based on twelve MMRd patients, these studies remain the largest cohorts in the literature (25, 32). In a literature review by Ran et al., 22 women with MMRd EC benefited from FST. Complete response at six months was estimated to be 47.8% and recurrence/progression was observed in 13%. Hysterectomy was performed in 34.8% of these patients. In this literature review, the differences between the two groups (Lynch syndrome and MLH1 hypermethylation) were not examined (16). The majority of these cases had Lynch Syndrome (nine patients in the study by Chung et al., seven patients in the study by Falcone et al., and three in the study by Ran et al.) (14, 16, 32). Only four patients, showing loss of expression of MLH1/PMS2 on immunohistochemistry consistent with an MLH1 promoter hypermethylation, were included in the study by Puechl et al. (15). Among these four patients, only one patient had progression and required hysterectomy.
Other studies have also hypothesized that MMRd responds poorly to hormone treatments usually used for FST (32, 33). Thus, fertility sparing seems inappropriate for this group of women and better subgroup refinement is needed.
POLE Mutated Endometrial Carcinomas: A Good Candidate for Fertility Sparing?
POLE mutated EC is the least common molecular subtype, accounting for approximately 6 to 10% of all ECs and 3% of ECs in young women (10, 14, 16, 19). POLE mutated EC has pathogenic mutations in the exonuclease domain of DNA polymerase epsilon, which is a protein involved in DNA replication. This results in extremely high somatic mutation frequencies (‘ultramutated’) exceeding 100 mutations per megabase.
It has been demonstrated that this subtype is particularly common in young women with high-grade endometrioid adenocarcinomas (grade 3 according to the WHO classification of the female tract tumors) and the presence of LVI (34). Despite these pathological features, POLE mutated EC is associated with a good prognosis and a reduced risk of recurrence and death (11, 35).
In a recent study on young women by Britton et al., POLE mutated EC was diagnosed at stage FIGO I in 85% of the cases and was frequently grade G1/G2. LVI was found in 18% of the cases with no nodal metastasis noted in their cohort. Furthermore, 25% of the cases had a myometrial infiltration >50%. In the group of women who underwent FST, only three had POLE mutated EC (16). Among these three women, only one had a progressive disease and two had hysterectomy (32). None of the patients had evidence of persistent disease (16). In the study of Dagher et al., only one patient had POLE mutated EC and had progression after progesterone treatment. She underwent hysterectomy and had no evidence of disease at the latest follow-up (25).
Indeed, in this population, the positive prognostic impact of the POLE mutation offers particularly relevant prospects for therapeutic de-escalation and justifies, according to the ESGO-ESTRO-ESP 2021 recommendations, monitoring for all stage I or II POLE mutated endometrioid EC after surgery (3-5). However, it remains unclear whether FST can be considered for this group, which has been frequently classified as high-grade tumors. However, the available data seem promising, although an extremely low number of patients has been reported in the literature (9). Unfortunately, there are no studies evaluating the impact of molecular classification on the treatment of young women with POLE mutated EC wishing to preserve fertility or the importance of the presence of LVI in this subtype, and there are no data available on their response to progestin treatments.
Conclusion
Fertility-sparing strategies for young patients with EC wishing to preserve fertility are becoming an important challenge for gynecological oncologists. New recommendations have emerged for certain patient subgroups. Molecular classification system plays an important role in predicting prognosis and has gained importance for therapeutic implications, including fertility preservation. Data are still extremely scarce in the literature, and approximately 100 patients have been described. We propose a patho-molecular-based algorithm for the selection of patients who may benefit from a fertility-sparing strategy (Figure 4) to better define this group of women. This algorithm differs from the currently recommended ProMisE algorithm because it considers the pathological grade, which is mandatory for FST. However, these recommendations should be applied with caution as the results of the clinical trials PORTEC4 and RAINBO are not yet available. Further clinical trials directed by molecular features are needed to refine the selection of patients eligible for fertility-sparing.
Algorithm for the selection of young patients with endometrial cancer wishing to preserve fertility according to the molecular subtype. *Stage IA was assessed using MRI and/or US. The absence of myometrial infiltration is mandatory. **The hysteroscopic specimen should allow the adequate pathological, immunohistochemical, and molecular assays. POLE: Polymerase epsilon; MMRd: mismatch repair deficiency; FSS: fertility-sparing strategy.
Footnotes
Authors’ Contributions
Conceptualization: CG and PM; methodology: MAB and PM; writing – original draft preparation: MAB; writing – review and editing, all Authors. All the Authors have read and agreed to the published version of the manuscript.
Conflicts of Interest
The Authors have no conflicts of interest to declare in relation to this study.
Funding
The Authors received no financial support for the research, authorship, or publication of this article.
- Received December 15, 2023.
- Revision received January 4, 2024.
- Accepted January 5, 2024.
- Copyright © 2024 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
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).