Secondary Logo

Journal Logo

Pathology of the Corpus: Original Articles

Molecular Evaluation of Low-grade Low-stage Endometrial Cancer With and Without Recurrence

Matrai, Cathleen E. M.D.; Ohara, Kentaro M.D., Ph.D.; Eng, Kenneth Wha M.S.; Glynn, Shannon M. M.D.; Chandra, Pooja M.D.; Chatterjee-Paer, Sudeshna M.D.; Motanagh, Samaneh M.D.; Mirabelli, Susanna B.S.; Kurtis, Boaz M.D.; He, Bing B.S.; Sigaras, Alexandros M.D.; Gupta, Divya M.D.; Chapman-Davis, Eloise M.D.; Holcomb, Kevin M.D.; Sboner, Andrea Ph.D.; Elemento, Olivier Ph.D.; Ellenson, Lora Hedrick M.D.; Mosquera, Juan Miguel M.D., M.Sc.

Author Information
International Journal of Gynecological Pathology: May 2022 - Volume 41 - Issue 3 - p 207-219
doi: 10.1097/PGP.0000000000000798
  • Open


Uterine cancer is the most common gynecologic cancer in developed countries and was the second most prevalent cancer among females in 2019 with over 60,000 new cases in the United States 1,2. It is estimated that by 2030, uterine cancer will be the third most common cancer among women, with a 2.9% average increase in incidence per year 3. Further, mortality rates have been increasing since 2000 4. Endometrial cancer (EC) is the most common type of uterine cancer, accounting for >90% of cases 1.

Despite these alarming statistics, the clinical approach to EC has remained largely stagnant, relying on a dualistic model developed over 30 yr ago 5. In this model, there are 2 defined subtypes of EC: type I, which is low grade, with endometrioid histology, hormone receptor positive and good prognosis, and type II, which is high grade, with nonendometrioid histology, often hormone receptor negative and has a poor prognosis 6. Most cases of EC are type 1 and can be defined as low-grade, low-stage endometrioid carcinomas (LGLS EC). These are treated primarily with surgery and the 5-yr survival rate is 95% 2. Current risk stratification systems utilize clinicopathologic information such as age, depth of myometrial invasion, histologic grade, and lymphovascular space involvement to define subgroups of women who are at high risk for recurrence 7–10. However, despite the development of multiple different stratification methods, predicting which patients with LGLS EC will eventually progress remains inaccurate and elusive 11,12. It is therefore imperative to determine the factors that predispose the small subgroup of patients with LGLS EC to behave more aggressively, which will inform the need for adjuvant therapy or more extensive initial intervention.

In the evolving age of molecular and genomic tumor characterization, there is promise in more precisely defining prognostic subtypes of EC. Data analyzed from the Cancer Genome Atlas (TCGA) revealed 4 distinct molecular subtypes of EC with different prognostic outcomes: microsatellite-instability, POLE ultra-mutated, copy-number low, and copy-number high 13. These findings have been replicated and demonstrate the immense potential for the use of molecular data in clinical risk stratification. Recent studies have suggested that the use of both molecular and clinicopathologic factors results in improved risk assessment and ability to predict outcomes 14–18.

Molecular analysis has uncovered a number of possible prognostic markers to help distinguish women in the LGLS subgroup who are at high risk of recurrence. It has been found that, among women with LGLS EC, mutations in CTNNB1 are associated with worse recurrence-free survival and overall survival 19–22. Further, women with CTNNB1-mutated tumors tend to have otherwise favorable characteristics such as young age, low tumor grade, less myometrial invasion and lower incidence of lymphatic and vascular invasion 21. In addition, a combination of high expression of NF-YA, a transcription factor involved in activating genes associated with growth promotion, and low expression of lamin-A, a protein associated with cell differentiation, have been noted as potential markers of tumor aggressiveness in LGLS EC 23. Low expression of ASRGL1 protein has also been found to have prognostic impact on survival in EC 24. Finally, while mutated in only a very small proportion of LGLS EC, aberrations in p53 are associated with worse recurrence-free survival 21, though this is a more complex issue. Among heterogeneous populations of EC with both low and high-grade tumors, it has been noted that TP53, along with PIK3CA, PTEN, and PPP2R1A, are likely early drivers of primary oncogenesis. However, among LGLS EC specifically, it remains unclear if mutations in TP53 are driving events in recurrence and metastasis or just synergistic with other molecular events underlying this process 25. Thus, the pathways involved in the process of tumor progression and recurrence in LGLS EC remain incompletely understood.

In this study, we analyzed the primary tumor and subsequent recurrence of LGLS ECs with the goal of determining molecular and genetic profiles of each, in the hopes of illuminating potential genetic risk factors for recurrence as well as pathways of progression.


Case Selection and Pathologic Examination

The study was performed under an institutional review board—approved protocol (WCM IRB# 1007011157). Twelve cases of LGLS endometrial endometrioid carcinoma cases were identified in our surgical pathology archives, including 4 cases with recurrence and 8 controls without recurrence matched for age, histology, and body mass index. Tumor morphology was reviewed by 2 pathologists with expertise in gynecologic pathology (L.H.E. and C.E.M.). For this study, low grade was defined as grade 1 and low stage was defined as stage 1a at diagnosis with comprehensive staging. Clinicopathologic data can be found in Table 1.

TABLE 1 - Clinicopathologic and molecular characteristics
Case Age (yr) Diagnosis FIGO stage Clinical characteristics Treatment Clinical follow-up Site of recurrence Time to recurrence Molecular findings
1 56 G1 EC Noninvasive 1A BMI: 27.8 No HRT Family history of ovarian cancer Initial: robotic total hysterectomy/BSO/LND Recurrence: surgical resection, chemoradiation also s/p subsequent adjuvant chemotherapy for a second primary peritoneal cancer NED after recurrence and second primary Midline pelvic mass ∼1 yr (12 mo) PTEN mut, PIK3CA mut, MYC mut*, 1q gain
2 69 G1 EC 30% invasive 1A BMI: 38.5 No HRT Initial: total abdominal hysterectomy/BSO/LND, vaginal cuff brachytherapy 1st Recurrence: surgical resection followed by radiation and antihormonal therapy 2nd Recurrence: surgical resection Lost to follow-up 2015 after second recurrence (1) Supraclavicular lymph node (2) Posterior neck mass 3 yr (∼37 mo) 7.5 yr (∼4.5 from first recurrence) PTEN mut, PIK3CA mut, ARID1A mut, ATM mut, TP53 mut, 8p deletion, 8q gain, 18p gain, 18q gain
3 54 G1 EC 12% invasive 1A BMI: 21.5 History of HRT Initial: robotic total hysterectomy/BSO/LND, adjuvant vaginal brachytherapy Recurrence: opted for no further therapy, inpatient hospice Died of disease 5/2013 Pelvic mass ∼5 yr (58 mo) PTEN mut, PIK3CA mut, ARID1A mut, EGFR mut
4 80 G1 EC 15% invasive 1A Obese (BMI unknown) No HRT Initial: total abdominal hysterectomy/BSO/LND/appendectomy Recurrence-pelvic radiation and vaginal cuff brachytherapy NED after recurrence Died of medical causes 2019 Pelvis/vagina ∼5.5 yr (67 mo) PTEN mut, PIK3CA mut, DDR2 mut, 1q gain
5 59 G1 EC Noninvasive 1A BMI: 34.5 No HRT Robotic total hysterectomy/BSO/LND NED for 9 yr (113 mo) after surgery None NA KIT mut
6 59 G1 EC 6% invasive 1A BMI: 28.3 No HRT Robotic total hysterectomy/BSO/LND NED at 2 yr (22 mo) then lost to follow-up None NA PTEN mut, PIK3CA mut, ARID1A mut
7 63 G1 EC Noninvasive 1A BMI: 25.8 No HRT Robotic total hysterectomy/BSO/LND NED for 11 yr (135 mo) after surgery None NA PIK3CA mut, ARID1A mut
8 67 G1 EC 33% invasive 1A BMI: 23 No HRT Robotic total hysterectomy/BSO/LND; declined vaginal brachytherapy NED for 9 yr (113 mo) after surgery None N/A KIT mut
9 53 G1 EC Noninvasive 1A BMI: 26.9 No HRT Robotic total hysterectomy/BSO/LND NED for 10 yr (122 mo) after surgery None NA PTEN mut, 19p gain, 21q gain
10 52 G1 EC Noninvasive 1A BMI: 30.5 No HRT Total laparoscopic hysterectomy/BSO/LND NED for 10 yr (120 mo) after surgery None NA PIK3CA mut, ARID1A mut, POLE mut, TMB high
11 79 G1 EC 7% invasive 1A BMI: 35 No HRT Robotic total hysterectomy/BSO/LND NED at 3 yr (38 mo) then lost to follow-up None NA PTEN mut, ARID1A mut
12 75 G1 EC 21% invasive 1A BMI: 20 No HRT Total laparoscopic hysterectomy/BSO/LND NED for 11 yr (133 mo) after surgery None NA PTEN mut, ARID1A mut
*Denotes findings identified in primary only.
Denotes findings identified in metastasis only.
BMI indicates body mass index; BSO, bilateral salpingo-oophorectomy; EC, endometrioid carcinoma; G1, grade 1; HRT, hormone replacement therapy; LND, lymph node dissection; mut, mutation; NA, not available; NED, no evidence of disease; OSH, outside hospital; PFS, progression-free survival; TMB, tumor mutational burden.

Whole-exome Sequencing (WES)

WES was performed on the patient tumor/matched germline DNA pair using Illumina HiSEquation 2500 Rapid Run Mode (2×101 bp; Illumina, San Diego, CA), as detailed in previously described protocols 26. Our clinical-grade WES test Exome Cancer Test Version 1 (EXaCT-1) has been approved by New York State Department of Health (ID# 43032), and has been described in detail in Rennert et al. 27. This approach allows for assessment of >21,000 genes, through the development and implementation of novel computational approaches for simultaneous detection of somatic point and indel mutations, copy-number variants, tumor mutational burden, and microsatellite instability (MSI) status. WES alterations were categorized based on their actionability and their clinical or biologic relevance. Alterations in 49 actionable or clinically significant genes were reported within “Category 1,” alterations in 508 known cancer-associated genes within “Category 2,” and somatic alterations of unknown significance within “Category 3.” Tumor mutation burden (TMB) was calculated for each sample as the number of mutations divided by the number of bases in the coverage space per million 26,28.

Calculating MSI by MSIsensor

MSI was detected by MSIsensor, a software tool that quantifies MSI in paired tumor-normal genome sequencing data and reports the somatic status of corresponding microsatellite sites in the human genome 29. MSIsensor score was calculated by dividing the number of microsatellite unstable by the total number of microsatellite stable sites detected.

Oncomine Comprehensive Assay

Previously reported tumor DNA and RNA extraction protocols and quantitation assays were followed 30. The Oncomine Comprehensive Assay version 3 (OCAv3) (ThermoFisher Scientific) was performed on the IonTorrent_S5 XL platform, following manufacturer protocols. Positive controls (Horizon Discovery) were used in each run. OCAv3 is an amplicon-based, targeted assay that enables the detection of relevant SNVs, amplifications, gene fusions, and indels from 161 unique genes (Table 2).

TABLE 2 - List of the genes covered by the oncomine comprehensive assay (OCAv3)
Hotspot genes Full length genes Copy number genes Gene fusions (intergenic and intragenic)

Detection of Somatic Copy Number Alterations (SCNA)

Copy number scores for matched tumor normal pairs in this cohort were calculated using CNVseeqer 31 as implemented in the EXaCT1 V0.9 pipeline 27. Capture regions with a total coverage <100 reads in both the tumor sample and matched control sample are filtered out. For SCNA, read counts are normalized in both the tumor sample and the matched control sample by the total number of reads aligned in the tumor sample and the matched control sample, respectively. Then the ratio of the normalized read counts in the tumor sample and the normalized read count in the control sample is calculated. These capture regions are then ordered karyotypically and sorted by genomic coordinates according to the log2 value of the ratio of normalized read counts of the tumor sample and control sample. The normalized ratios of these bins were segmented using the Circular Binary Segmentation algorithm implemented in the R package DNAcopy 32. The algorithm outputs segments where every capture region found within these segments is represented by the same log2 value. This log2 value indicates whether the segment has DNA copy number gain (amplification) or DNA copy number loss (deletion). A negative log2 would suggest a segment was deleted and a positive value would suggest a segment is amplified. Segments with a log2 value >0.5 to are amplified and segments with a log2 value <−0.5 are categorized as deleted. We then took the segments called by the algorithm and with a custom script annotated these segments by RefSeq genes whose transcription start and end sites overlap with the genomic coordinates assigned to these segments. Copy number alteration (CNA) burden was calculated as the percentage of the cancer genome showing copy number changes according to a previously reported method 33.

Correcting SCNA for Tumor Purity

The copy number scores were then adjusted for ploidy and purity corrections using CLONET 34. In tumor samples where purity could not be estimated by CLONET, we used pathology estimates of tumor purity. Copy number scores that could not be adjusted by CLONET were adjusted by the following formula

Computing SCNA by Chromosomal Arm

We binned the hg19/b37 genome into chromosomal arms. Genomic coordinates for the p (short) arm and q (long) arm, for each chromosome is obtained using UCSC cytoband file ( Using the segment calls generated by CNVseeqer, log2 scores associated with each segment were corrected using the previously described method above. The corrected segmented calls were binned into the chromosomal arm based on the genomic coordinates using a custom script. To compute the SCNA burden for each arm, the fraction of bases altered was calculated based on log2 threshold (>0.5 for amplification and <−0.5 for deletion).

Sanger Sequencing

For genes of interest, PCR was performed with appropriate human gDNA control samples using custom PCR primers designed to amplify short (~200–400 bp) regions in FFPE samples, as previously described 30. Data analysis was performed with DNASTAR Lasergene12 software and the threshold for SNP detection was set to 10%. Mutations from the reference sequence were called whenever sequence quality and coverage allowed.


Summary of Clinical Characteristics

Detailed clinicopathologic characteristics are summarized in Table 1. Four cases of LGLS EC with recurrence (including primary tumor and recurrence) and 8 cases without recurrence were identified. Sites of recurrence included the pelvis, vagina, lymph node, and neck. One patient (case 2) received adjuvant vaginal brachytherapy following their initial surgery (performed prior to updated 2009 FIGO staging for EC). Follow-up was available for all patients with a mean of 114 mo (range: 24–180 mo). An average observational period for cases of LGLS EC without recurrence was 100 mo (range: 24–135 mo). All patients were postmenopausal with an average age of 67 yr.


All available hematoxylin and eosin (H&E) slides from each case were reviewed by 2 pathologists with expertise in Gynecologic Pathology (L.H.E. and C.E.M.). Cases were assessed for classic morphology and all diagnoses were confirmed. All cases displayed characteristic features of well-differentiated grade 1 endometrioid carcinoma, including glandular configuration, lack of significant solid growth, smooth luminal borders, fencepost nuclei, and mild to moderate nuclear atypia at most. Case 4 showed focal mucinous differentiation. Histologic features can be seen in Figures 1 and 2. Although the cytologic atypia in the recurrence for case 2 was slightly higher than that in the primary, all the recurrent tumors also displayed low-grade architectural and cytologic features.

FIG. 1:
Histopathology of low-grade low-stage endometrioid carcinoma with recurrence. Top row corresponds to primary tumors and bottom row corresponds to recurrences, represented by low and high (insets) magnification images. (A) Noninvasive well-differentiated tumor (case 1) and (D) recurrence in pelvis, 2 yr later. Sanger sequencing confirmed PIK3CA and FGFR2 mutations. (B) Invasive (30% myometrial thickness) well-differentiated tumor (case 2) and (E) recurrence in right posterior neck (frozen material), 8 yr later. Sanger sequencing confirmed PTEN and TP53 mutations. (C) Invasive (15% myometrial thickness) well-differentiated tumor with mucinous features (case 4) and (F) recurrence in vagina, 6 yr later. Sanger sequencing confirmed KRAS and MAX mutations.
FIG. 2:
Histopathology of low-grade low-stage endometrioid carcinoma without recurrence. (A) Noninvasive, well-differentiated tumor (case 5). Sanger sequencing confirmed AKT1 mutation. (B) Invasive (33% of myometrium) well-differentiated tumor (case 8). Sanger sequencing confirmed CCND1 mutation.

Genomic Landscape of LGLS EC With Recurrence and Without Recurrence

The 12 primary LGLS EC underwent WES and Oncomine, and 3 of 4 recurrent tumors were interrogated by Oncomine (Fig. 3). The median coverage was 89.8× and 91.2×, respectively, by WES for tumor and germline samples and 1362× by OCAv3. Exclusion of the fourth case was based upon insufficient material. Two of the 3 recurrent cases showed additional mutations in the recurrence.

FIG. 3:
Genomic landscape of low-grade low-stage endometrioid carcinoma (LGLS EC) with and without recurrence. (A) The oncoprint shows cancer genes that were altered by mutation and/or copy number change in at least 2 tumor samples, except for TP53. Mutations of unknown significance are not included. Each column corresponds to 1 tumor sample. (B and C) MSIsensor and TMB scores in LGLS EC with and without recurrence. A blue dashed line represents a threshold for microsatellite instability. Black lines represent averages of the scores in each group. NS indicates not significant; TMB, tumor mutational burden; WES, whole-exome sequencing.

The most commonly observed genomic mutations were PTEN (73.3%), PIK3CA (66.7%), and ARID1A (53.3%), which were consistent with previously reported results from TCGA 13. We found no significant difference in the prevalence of mutations between LGLS EC with recurrence and without recurrence. Comparing the mutations between primary LGLS EC and the matched recurrent tumors, we identified an additional loss-of-function TP53 mutation (p.Gly245Asp) in case 2 and POLE nonsense mutation (p.Glu1951Ter) and a receptor tyrosine kinase DDR2 mutation (p.Leu239Met) in case 4. The identified POLE mutation occurred outside the proofreading domain (residue 86–426). The DDR2 mutation is predicted to be damaging by PolyPhen-2 35. The recurrent tumor from case 1 did not accrue additional mutations. Case 3 recurrence had insufficient material for sequencing. PIK3CA mutations were detected in 4 of 4 primary LGLS EC with recurrence and in 3 of 8 disease-free cases.

Secondly, we examined WES data from primary tumors to determine whether MSI status and TMB of LGLS EC with recurrence differ from those without recurrence. Three LGLS EC with recurrence and 2 without recurrence were MSI high (Fig. 3B). MSIsensor scores in LGLS EC with recurrence were significantly higher than those without recurrence. TMB in one LGLS EC without recurrence, which harbored a POLE mutation (p.Ser1380Leu, case 10), was notably high (Fig. 3C). There was no statistically significant difference in TMB between LGLS EC with recurrence and without recurrence.

We found no germline mutations in MSI-related or cancer-related genes in cases with LGLS EC with recurrence, while AKT1 variants (p.Glu17Lys) were identified in case 5 and APC nonsense mutation (p.Arg876*) in case 10.

Somatic Copy Number Alterations in LGLS EC With Recurrence and Without Recurrence

SCNAs were assessed for all primary LGLS EC using WES data. The SCNA landscape revealed arm-level broad copy number alterations in 3 LGLS EC with recurrence (Fig. 4A). We observed 1q gain in 2 LGLS EC with recurrence (cases 1 and 4), and 8p deletion, 8q gain, 18p, and 18q gain in 1 case with recurrence (case 2). Arm-level CNA analysis confirmed these alterations and additionally revealed arm-level alterations in 19p and 21q in 1 LGLS EC without recurrence (case 9). CNA burden, which represents the level of copy number gains and losses across the genome was similar between LGLS EC with and without recurrence (Fig. 4B). When CNA burden was restricted to regions showing gains, CNA burden in LGLC EC with recurrence was significantly higher than those without recurrence (Fig. 4C).

FIG. 4:
Comparison of somatic copy number alterations (SCNA) between tumors with and without recurrence. (A) Overall picture of SCNA in low-grade low-stage endometrioid carcinoma (LGLS EC) with and without recurrence based on whole-exome sequencing data; dark blue, homologous deletion; light blue, heterologous deletion; pink, gain; red, amplification. Each row corresponds to 1 LGLS EC. Dashed lines represent borders between short arms and long arms in each chromosome. (B) Comparison of CNA burden. CNA burden represents the level of copy number gain and loss. Black lines represent averages of the scores in each group. (C) Comparison of CNA burden values, which are restricted to copy number gain. Black lines represent averages of the scores in each group. CAN indicates copy number alteration; NS, not significant.

Sanger Sequencing

Sanger sequencing confirmed a selection of mutations in representative cases: PIK3CA in case 1 primary, FGFR2 in case 1 recurrence, PTEN in case 2 primary, TP53 in case 2 recurrence, KRAS in case 4 primary and MAX in case 4 recurrence, AKT1 on case 5.


Although LGLS EC has a 95% 5 yr survival, those patients who recur have a poor prognosis and limited treatment options 5,19,36,37. The underlying cause as to why certain tumors recur despite a lack of histologic or clinical evidence portending aggressive behavior has not yet been elucidated. Our data support that a subset of LGLS EC may be genetically predisposed to recurrence. In this pilot study, 3 of 4 recurrent cases showed segments affected by copy number gain or amplification and the MSIsensor scores were significantly higher than non-recurrent cases, with the exception of 1 case with an extremely high tumor mutation rate and POLE mutation (801 counts of nonsynonyous mutations, case 10). Recurrent cases also showed higher TMB burden than nonrecurrent cases, again with the exception of case 10. All recurrent cases also showed PIK3CA mutation (Fig. 1) while only 3 of 8 controls did. In addition, 2 of 3 recurrent cases gained a mutation associated with genetic instability (TP53 and POLE) and 1 case acquired a mutation in DDR2, a gene encoding a receptor tyrosine kinase which is a potential therapeutic target, results which may have important clinical implications. A prior in vivo study suggests that DDR2 mutations are a potential therapeutic target in lung squamous cell carcinoma 33. Another recent study has shown that targeting DDR2 kinase increases the efficacy of anti-PD-L1 therapy 38.

The TCGA has noted that 25% of high-grade endometrioid tumors and a limited percentage of lower grade (1 and 2) endometrioid tumors (<5%) show extensive copy number alterations. By WES analysis, the recurrent primaries in our study showed higher levels of copy number gains than controls. This finding is of clinical importance because a previous report suggested that high SCNA burden is a prognostic factor in TCGA cohorts, including EC cases 36. Further investigation into the significance of these findings in larger cohorts would be warranted to identify patients who may be at higher risk of recurrence and need alternative treatment strategies.

SCNA analysis also demonstrated that arm-level 1q gain was enriched in LGLS cases with recurrence although the trend did not reach statistical significance. 1q gain has been documented in association with worse prognosis in a subset of tumor types including multiple myeloma, Ewing sarcoma, pediatric tumors including Wilms tumor and neuroblastoma, and multiple brain tumors including ependymoma and high-grade gliomas 39–43. To our knowledge there are only a few reports in the literature addressing gains in 1q in the endometrial carcinoma. Depreeuw et al. 44 reported that 1q32.1 amplification significantly correlated with worse relapse-free survival and MDM4 is a potential oncogenic driver in the region in a nonspecific molecular profile subgroup of EC. Of note, MDM4 gain was detected by WES in 2 of 4 LGLS EC with recurrence (Fig. 3A). Sever et al. 45 also reported that gains in 1q had a negative impact on survival in EC patients. In addition, aberrations in 1q have also been described in Müllerian tumors of mesonephric type, though neither of our cases showed histologic features of this tumor type, despite the fact that case 4 also harbored KRAS mutation, a finding also documented in a high percentage of mesonephric carcinomas 46–48.

Tumor mutational burden was lower in the cases with recurrence (Fig. 3C; average total number of nonsynonymous mutations=118 vs. 130), though not to statistical significance.

The same held true for the TMB density (3.19 vs. 3.5 mut/Mb). The difference was reversed when adjusting for nonrecurrent case 10 (TMB=118 vs. 33, P=0.024 2-sided; TMB density=3.19 vs. 0.9 mut/Mb, P=0.024), which showed a markedly higher total number of mutations (801) and TMB density (21.7 mut/Mb) in comparison to the remaining control cases.

Importantly, POLE alteration was identified in this case on WES, which explains the high TMB without associated impact on outcome. This patient was recurrence free at last follow up, which would be in keeping with the proposed improved prognosis associated with POLE mutated tumors 49–52. It is unclear whether the POLE mutation detected in case 10 (p.Ser1380Leu) is pathogenic because the mutation is not located in the exonuclease domain. However, 1 ultramutated tumor in TCGA EC cohort harbored the same POLE mutation as case 10 13, which may provide evidence that the observed POLE mutation could lead to extensive somatic mutations.

A notable absence in our data is that of CTNNB1 mutations, which have been shown to be associated with more aggressive behavior in carcinomas of endometrioid histology with higher rates of recurrence and decreased survival 19,21,22. Kurnit and colleagues found that 26% of LGLS endometrioid patients harbored mutation in CTNNB1, though they defined this as grades 1 and 2 and stages 1 and 2, while Moroney and colleagues found overall 60% of LGLS (defined as grade 1 stage 1) showed CTNNB1 mutation (vs. 28% of controls). None of the 4 recurrent cases in our study showed alteration in CTNNB1 while 2 of 8 nonrecurrent cases did. The small sample size in the current study may account for this discrepancy. In addition, given tumoral heterogeneity seen in LGLS cases 53, it is possible that sampling could play a role, though it would be unusual to occur in all 4 cases.

Although there have been reports of low and intermediate grade endometrioid carcinomas with TP53 mutation and associated reduced survival 21,25, these tumors appear to be quite uncommon, particularly with respect to grade 1 tumors. Interestingly, we identified a TP53 mutation (p.Gly245Asp) in recurrent case 2 (Fig. 1E). Kurnit and colleagues showed an 8.8% incidence of TP53 mutations when grouping grade 1 and 2 tumors, and the TCGA study showed 0% and 11.8% for grade 1 and 2 tumors, respectively 21,49. When present in endometrioid carcinomas, TP53 mutations are overwhelmingly associated with higher grade morphology 16. In addition, TP53 mutations may occur in association with other markers of genetic instability (MMR/POLE), and the clinical impact of these “passenger mutations” may be minimal, though they still remain a challenge in terms of accurate classification. It is generally accepted that the incidence of TP53 mutations in low-grade tumors is very low, a finding corroborated in this study, including in cases demonstrating subsequent recurrence.

Among nonrecurrent cases, 3 of 8 had KIT mutations while 1 was seen in 1 of 4 recurrent cases (Fig. 3A). In the setting of gynecologic malignancy, the primary focus of KIT mutation in recent years has been in uterine sarcomas/MMMT 54–56. While data regarding other EC subtypes is limited, a few studies have looked at the incidence of KIT mutation/expression in endometrioid ECs, with numbers ranging from 0% (immunohistochemistry and molecular) 57 to 25% to 60% (immunohistochemistry only) 58,59. Slomovitz and colleagues showed 2 of 8 recurrent endometrioid tumors stained positively for c-kit while 0 of 33 nonrecurrent cases did. Other researchers found that EC cases with C-KIT immunohistochemical expression more frequently had metastases and shorter disease-free survival 59. Still, the majority of these studies looked at immunohistochemical expression rather than sequencing data. Kafshdooz et al. 60 proposed that the pattern and frequency of KIT mutation differed between tumors of different stages, but they did not comment on the specifics of incidence in the groups. Hence, data regarding KIT mutation in carcinomas of the endometrium remains somewhat limited. All 3 recurrences available for sequencing showed shared mutations with the primary. Two gained mutations associated with genetic instability (TP53 and POLE, cases 2 and 4, respectively) and case 4 also acquired a mutation in DDR2 kinase, a potential therapeutic target. Case 1 showed no additional mutations and lacked the MYC mutation seen in the primary tumor (Fig. 3A). Few studies have investigated the mutation profiles related to progression of EC from primary to metastatic lesions. Gibson and colleagues analyzed complex atypical hyperplasia, primary tumors, and paired abdominopelvic metastatic lesions of EC of both endometrioid and nonendometrioid histologies with WES and found heterogeneity among all stages, with only about 50% of mutations being shared between metastatic lesions and primary tumors. Further, no recurrent, metastasis-specific mutations were found 61. As LGLS EC is a heterogenous subgroup itself, there is a need to study this population individually to better characterize molecular events related to tumor progression.

It is unclear whether the POLE and DDR2 mutations additionally detected in the recurrent tumor from Case 4 promoted tumor recurrence. POLE-ultramutated ECs are defined by the presence of a somatic mutation in the exonuclease domain of POLE gene 62. The detected nonsense mutation does not result in an amino acid change in the exonuclease domain, which suggests that the proofreading activity of POLE is preserved. In addition, targeted sequencing did not show very high mutation frequency in the recurrent tumor from case 4 (6 nonsynonymous mutations). On the other hand, the DDR2 mutation was predicted to be functionally damaging by a computational tool for prediction of the possible impact of an amino acid substitution. While in vitro studies using tumor cell lines have proven DDR2 protein exerts oncogenic role by promoting cellular proliferation, migration and metastasis 63, the oncogenic role resulting from the DDR2 mutation is unclear in EC. Further functional studies would be warranted in order to clarify whether the DDR2 mutation promotes malignant progression of ECs.

The underlying cause as to why certain LGLS ECs recur despite a lack of histologic or clinical evidence portending aggressive behavior has not yet been elucidated. The increasing use of molecular studies in characterizing EC has resulted in notable progress in further defining these tumors. As molecular testing continues to become more commonplace in practice, it is imperative that we continue to seek out molecular identifiers for patients who are at risk for recurrence, as there is untapped potential to preemptively identify patients who are at risk for recurrent and typically untreatable disease. Although this is a small WES-based and targeted sequencing study, it supports that molecular differences may define cases of LGLS EC that recur and those that do not.


1. Miller KD, Nogueira L, Mariotto AB, et al. Cancer treatment and survivorship statistics, 2019. CA Cancer J Clin 2019;69:363–85.
2. Ferlay J, Soerjomataram I, Dikshit R, et al. Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer 2015;136:E359–86.
3. Rahib L, Smith BD, Aizenberg R, et al. Projecting cancer incidence and deaths to 2030: the unexpected burden of thyroid, liver, and pancreas cancers in the united states. Cancer Res 2014;74:2913–21.
4. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2016. CA Cancer J Clin 2016;66:7–30.
5. McAlpine JN, Temkin SM, Mackay HJ. Endometrial cancer: not your grandmother’s cancer. Cancer 2016;122:2787–98.
6. Bokhman JV. Two pathogenetic types of endometrial carcinoma. Gynecol Oncol 1983;15:10–7.
7. Creutzberg CL, Van Putten WLJ, Koper PCM, et al. Surgery and postoperative radiotherapy versus surgery alone for patients with stage-1 endometrial carcinoma: multicentre randomised trial. Lancet 2000;355:1404–11.
8. Keys HM, Roberts JA, Brunetto VL, et al. A phase III trial of surgery with or without adjunctive external pelvic radiation therapy in intermediate risk endometrial adenocarcinoma: a Gynecologic Oncology Group study. Gynecol Oncol 2004;92:744–51.
9. Colombo N, Creutzberg C, Amant F, et al. ESMO-ESGO-ESTRO consensus conference on endometrial cancer: diagnosis, treatment and follow-up. Radiother Oncol 2015;117:559–81.
10. Todo Y, Kato H, Kaneuchi M, et al. Survival effect of para-aortic lymphadenectomy in endometrial cancer (SEPAL study): a retrospective cohort analysis. Lancet 2010;375:1165–72.
11. Bendifallah S, Canlorbe G, Collinet P, et al. Just how accurate are the major risk stratification systems for early-stage endometrial cancer? Br J Cancer 2015;112:793–801.
12. Liu Y, Broaddus RR, Zhang W. Identifying aggressive forms of endometrioid-type endometrial cancer: new insights into molecular subtyping. Expert Rev Anticancer Ther 2014;15:1–3.
13. Getz G, Gabriel SB, Cibulskis K, et al. Integrated genomic characterization of endometrial carcinoma. Nature 2013;497:67–73.
14. Talhouk A, Mcconechy MK, Leung S, et al. A clinically applicable molecular-based classification for endometrial cancers. Br J Cancer 2015;113:299–310.
15. Talhouk A, McAlpine JN. New classification of endometrial cancers: the development and potential applications of genomic-based classification in research and clinical care. Gynecol Oncol Res Pract 2016;3:14.
16. Talhouk A, McConechy MK, Leung S, et al. Confirmation of ProMisE: a simple, genomics-based clinical classifier for endometrial cancer. Cancer 2017;123:802–13.
17. Stelloo E, Nout RA, Osse EM, et al. Biology of human tumors improved risk assessment by integrating molecular and clinicopathological factors in early-stage endometrial cancer—combined analysis of the PORTEC Cohorts. Clin Cancer Res 2016;22:4215–24.
18. Talhouk A, Derocher H, Schmidt P, et al. Molecular subtype not immune response drives outcomes in endometrial carcinoma. Clin Cancer Res 2019;25:2537–48.
19. Moroney MR, Davies KD, Wilberger AC, et al. Molecular markers in recurrent stage I, grade 1 endometrioid endometrial cancers. Gynecol Oncol 2019;153:517–20.
20. Myers A, Barry WT, Hirsch MS, et al. β-Catenin mutations in recurrent FIGO IA grade I endometrioid endometrial cancers. Gynecol Oncol 2014;134:426–7.
21. Kurnit KC, Kim GN, Fellman BM, et al. CTNNB1 (beta-catenin) mutation identifies low grade, early stage endometrial cancer patients at increased risk of recurrence. Mod Pathol 2017;30:1032–41.
22. Liu Y, Patel L, Mills GB, et al. Clinical significance of CTNNB1 mutation and Wnt pathway activation in endometrioid endometrial carcinoma. J Natl Cancer Inst 2014;106:dju245.
23. Cicchillitti L, Corrado G, Carosi M, et al. Prognostic role of NF-YA splicing isoforms and Lamin A status in low grade endometrial cancer. Oncotarget 2017;8:7935–45.
24. Fonnes T, Berg HF, Bredholt T, et al. Asparaginase-like protein 1 is an independent prognostic marker in primary endometrial cancer, and is frequently lost in metastatic lesions. Gynecol Oncol 2018;148:197–203.
25. Fadare O, Parkash V. p53 aberrations in low grade endometrioid carcinoma of the endometrium with nodal metastases: possible insights on pathogenesis discerned from immunohistochemistry. Diagn Pathol 2017;12:81.
26. Beltran H, Eng K, Mosquera JM, et al. Whole-exome sequencing of metastatic cancer and biomarkers of treatment response. JAMA Oncol 2015;1:466–74.
27. Rennert H, Eng K, Zhang T, et al. Development and validation of a whole-exome sequencing test for simultaneous detection of point mutations, indels and copy-number alterations for precision cancer care. NPJ Genomic Med 2016;1:16019.
28. Sailer V, Eng KW, Zhang T, et al. Integrative molecular analysis of patients with advanced and metastatic cancer. JCO Precis Oncol 2019;3:PO.19.00047.
29. Niu B, Ye K, Zhang Q, et al. MSIsensor: microsatellite instability detection using paired tumor-normal sequence data. Bioinformatics 2014;30:1015–6.
30. Ginter PS, McIntire PJ, Kurtis B, et al. Adenomyoepithelial tumors of the breast: molecular underpinnings of a rare entity. Mod Pathol 2020;33:1764–72.
31. Jiang Y, Redmond D, Nie K, et al. Deep sequencing reveals clonal evolution patterns and mutation events associated with relapse in B-cell lymphomas. Genome Biol 2014;15:432.
32. Olshen AB, Venkatraman ES, Lucito R, et al. Circular binary segmentation for the analysis of array-based DNA copy number data. Biostatistics 2004;5:557–72.
33. Hammerman PS, Sos ML, Ramos AH, et al. Mutations in the DDR2 kinase gene identify a novel therapeutic target in squamous cell lung cancer. Cancer Discov 2011;1:78–89.
34. Prandi D, Baca SC, Romanel A, et al. Unraveling the clonal hierarchy of somatic genomic aberrations. Genome Biol 2014;15:439.
35. Adzhubei IA, Schmidt S, Peshkin L, et al. A method and server for predicting damaging missense mutations. Nat Methods 2010;7:248–9.
36. Hieronymus H, Schultz N, Gopalan A, et al. Copy number alteration burden predicts prostate cancer relapse. Proc Natl Acad Sci U S A 2014;111:11139–44.
37. Creasman WT, Odicino F, Maisonneuve P, et al. Carcinoma of the Corpus Uteri. Int J Gynaecol Obstet 2006;95(suppl 1):S105–S143.
38. Tu MM, Lee FYF, Jones RT, et al. Targeting DDR2 enhances tumor response to anti-PD-1 immunotherapy. Sci Adv 2019;5:eaav2437.
39. Rajeshwari M, Sharma MC, Kakkar A, et al. Evaluation of chromosome 1q gain in intracranial ependymomas. J Neurooncol 2016;127:271–8.
40. Shah GL, Landau H, Londono D, et al. Gain of chromosome 1q portends worse prognosis in multiple myeloma despite novel agent-based induction regimens and autologous transplantation. Leuk Lymphoma 2017;58:1823–31.
41. Schmidt TM, Barwick BG, Joseph N, et al. Gain of chromosome 1q is associated with early progression in multiple myeloma patients treated with lenalidomide, bortezomib, and dexamethasone. Blood Cancer J 2019;9:1–13.
42. Warwick AB, Dome JS. Lanzkowsky’s Manual of Pediatric Hematology and Oncology, 6th Edition. London, UK: Academic press; 2016;25:491–504.
43. Puri L, Saba J. Getting a clue from 1q: gain of chromosome 1q in cancer. J Cancer Biol Res 2014;2:1053.
44. Depreeuw J, Stelloo E, Osse EM, et al. Amplification of 1q32.1 refines the molecular classification of endometrial carcinoma. Clin Cancer Res 2017;23:7232–41.
45. Sever E, Doğer E, Çakiroğlu Y, et al. The effect of first chromosome long arm duplication on survival of endometrial carcinoma. Turk J Obstet Gynecol 2014;11:207–10.
46. Montalvo N, Redrobán L, Galarza D. Mesonephric adenocarcinoma of the cervix: a case report with a three-year follow-up, lung metastases, and next-generation sequencing analysis. Diagn Pathol 2019;14:71.
47. Cavalcanti MS, Schultheis AM, Ho C, et al. Mixed mesonephric adenocarcinoma and high-grade neuroendocrine carcinoma of the uterine cervix: case description of a previously unreported entity with insights into its molecular pathogenesis. Int J Gynecol Pathol 2017;36:76–89.
48. Mirkovic J, Sholl LM, Garcia E, et al. Targeted genomic profiling reveals recurrent KRAS mutations and gain of chromosome 1q in mesonephric carcinomas of the female genital tract. Mod Pathol 2015;28:1504–14.
49. Levine D, The Cancer Genome Atlas Research Network. Integrated genomic characterization of endometrial carcinoma. Nature 2013;497:67–73.
50. Stelloo E, Bosse T, Nout RA, et al. Refining prognosis and identifying targetable pathways for high-risk endometrial cancer; a TransPORTEC initiative. Mod Pathol 2015;28:836–44.
51. Church DN, Stelloo E, Nout RA, et al. Prognostic significance of POLE proofreading mutations in endometrial cancer. J Natl Cancer Inst 2014;107:402. doi: 10.1093/jnci/dju402
52. van Gool IC, Eggink FA, Freeman-Mills L, et al. POLE proofreading mutations elicit an antitumor immune response in endometrial cancer. Clin Cancer Res 2015;21:3347–55.
53. Lazo de la Vega L, Samaha MC, Hu K, et al. Multiclonality and marked branched evolution of low-grade endometrioid endometrial carcinoma. Mol Cancer Res 2019;17:731–40.
54. Rushing RS, Shajahan S, Chendil D, et al. Uterine sarcomas express KIT protein but lack mutation(s) in exon 11 or 17 of c-KIT. Gynecol Oncol 2003;91:9–14.
55. Raspollini MR, Pinzani P, Simi L, et al. Uterine leiomyosarcomas express KIT protein but lack mutation(s) in exon 9 of c-KIT. Gynecol Oncol 2005;98:334–5.
56. Lee CH, Hoang LN, Yip S, et al. Frequent expression of KIT in endometrial stromal sarcoma with YWHAE genetic rearrangement. Mod Pathol 2014;27:751–7.
57. Vandenput I, Debiec-Rychter M, Capoen A, et al. Kit gene in endometrial carcinoma: an immunohistochemical and mutational analysis. Int J Gynecol Cancer 2011;21:203–5.
58. Slomovitz BM, Broaddus RR, Schmandt R, et al. Expression of imatinib mesylate-targeted kinases in endometrial carcinoma. Gynecol Oncol 2004;95:32–6.
59. Scobie JV, Acs G, Bandera CA, et al. C-kit immunoreactivity in endometrial adenocarcinomas and its clinicopathologic significance. Int J Gynecol Pathol 2003;22:149–55.
60. Kafshdooz T, Ardabili SMM, Kafshdooz L, et al. C-kit mutations in endometrial cancer: Correlation with tumor histologic type. Asian Pacific J Cancer Prev 2015;16:7449–52.
61. Gibson WJ, Hoivik EA, Halle MK, et al. The genomic landscape and evolution of endometrial carcinoma progression and abdominopelvic metastasis. Nat Genet 2016;48:848–55.
62. Rayner E, van Gool IC, Palles C, et al. A panoply of errors: polymerase proofreading domain mutations in cancer. Nat Rev Cancer 2016;16:71–81.
63. Valiathan RR, Marco M, Leitinger B, et al. Discoidin domain receptor tyrosine kinases: new players in cancer progression. Cancer Metastasis Rev 2012;31:295–321.

Low-grade low-stage endometrioid carcinoma; Recurrence; Whole exome sequencing; Microsatellite instability score; Copy number analysis

Copyright © 2021 The Author(s). Published by Wolters Kluwer Health, Inc.