Mesonephric-like Adenocarcinoma of the Uterine Corpus: Genomic and Immunohistochemical Profiling with Comprehensive Clinicopathological Analysis of 17 Consecutive Cases from a Single Institution

Data on genetic and immunophenotypical characteristics of uterine mesonephric-like adenocarcinoma (MLA) remain limited. Therefore, we aimed to investigate the clinicopathological, immunohistochemical, and molecular features of uterine MLA. We performed targeted sequencing, array comparative genomic hybridization, and immunostaining in 17, 13, and 17 uterine MLA cases, respectively. Nine patients developed lung metastases. Eleven patients experienced disease recurrences. The most frequently mutated gene was Kirsten rat sarcoma viral oncogene homolog (KRAS; 13/17). Both the primary and matched metastatic tumors harbored identical KRAS (3/4) and phosphatase and tensin homolog deleted on chromosome 10 (1/4) mutations, and did not harbor any additional mutations. A total of 2 of the 17 cases harbored tumor protein 53 (TP53) frameshift insertion and deletion, respectively. Chromosomal gains were detected in 1q (13/13), 10 (13/13), 20 (10/13), 2 (9/13), and 12 (6/13). Programmed cell death-ligand 1 overexpression or mismatch repair deficiency was not observed in any of the cases. Initial serosal extension and lung metastasis independently predicted recurrence-free survival with hazard ratios of 6.30 and 7.31, respectively. Our observations consolidated the clinicopathological and molecular characteristics of uterine MLA. Both clinicians and pathologists should consider these features to make an accurate diagnosis of uterine MLA and to ensure appropriate therapeutic management of this rare entity.


Introduction
Endometrial carcinoma (EC) is the sixth leading cause of carcinoma-related death among women worldwide [1][2][3].The incidence rates of EC have been steadily increasing, particularly in developed countries [4,5].The diagnosis of EC subtypes is based on their distinct morphological features [6].Endometrioid carcinoma accounts for the majority of EC cases, followed by serous carcinoma, clear cell carcinoma, carcinosarcoma, and undifferentiated carcinoma [6].The histological type and grade as well as the International Federation of Gynecology and Obstetrics (FIGO) staging guides EC prognosis [6,7]; however, histological features overlap significantly between some subtypes, which makes accurate classification difficult.Consequently, continued efforts have been made to develop ancillary techniques, such as immunohistochemical staining (IHC) and molecular analyses, to stratify EC patients.
Mesonephric-like adenocarcinoma (MLA) is a rare but distinct gynecological malignancy, primarily arising in the uterine corpus and adnexa.MLA was recently introduced as a new histological type in the 2020 version of the World Health Organization Classification of Female Genital Tumors [6].Uterine MLA exhibits morphological, immunophenotypical, and genetic characteristics similar to those of mesonephric adenocarcinoma (MA), which is a rare malignant tumor derived from the mesonephric remnants located in the lateral wall of the vagina and uterine cervix [8].The term MLA has been used for malignant mesonephric lesions arising in the uterine corpus [3,9,10], because it still debated whether they are of mesonephric origin.Both clinicians and pathologists should be aware of the clinical, pathological, and molecular features of uterine MLA to make an accurate diagnosis and to ensure appropriate therapeutic management.
Next-generation sequencing (NGS), also known as massively parallel sequencing, allows for the effective capture of a substantial amount of genomic information regarding tumor development, progression, and biological behavior [11].NGS is inextricably intertwined with the realization of precision medicine in oncology.While it is unlikely to obviate traditional pathological diagnosis in its current state, NGS allows a more complete picture of carcinogenesis, progression, and metastasis than can be seen with any other modality.Recent advances in sequencing technologies have provided substantial insights into the mutated carcinoma-related genes and mutational processes operative in EC [12,13].Novel classification systems that incorporate molecular features have been developed to provide objective and reproducible EC categorization [14].The addition of molecular and immunohistochemical markers allows for the identification of clinically relevant subgroups, with potential therapeutic implications [15].However, despite significant advances directed toward elucidating molecular mechanisms and developing clinical trials for patients with EC, data on specific genetic alterations and molecular characteristics of uterine MLA remain limited [3,16,17].
Immune checkpoints and inhibitory immunoreceptors, including programmed cell death protein 1 (PD-1) and its ligand (PD-L1), have gained attention as therapeutic targets [18].PD-L1 expression on tumor and immune cells can be detected using IHC and different PD-L1 commercial clones.PD-1 and PD-L1 play progressively important roles in our understanding of tumor immunology and antitumor treatment [19].Binding of PD-L1 to its receptor PD-1 leads to T-cell inactivation in a variety of carcinomas [20].Therefore, anti-PD-1/PD-L1 treatment deregulates the adverse impact of tumor-infiltrating T-cells, which in turn may reverse the tumor immune resistance [19].Several clinical and experimental studies have investigated PD-L1 expression in EC and its prognostic values as well as its efficacy as an immunotherapy for EC [21][22][23]; however, these studies focused on endometrioid carcinomas and did not consider MLA as a separate or independent group.In addition, clinical trials that investigated the anti-PD-1 antibody, pembrolizumab, as a treatment for advanced or recurrent EC [24] did not separate MLA from other histological EC types.
Our group has steadily documented the clinical manifestations, cytological and histological features, immunophenotypes, and mutational profiles of uterine MLA using IHC and molecular analyses [2,3,[33][34][35][36][37][38][39][40].However, get a deep insight into their relevance, the results need to be consolidated and completely analyzed.Therefore, in this study, we comprehensively investigated the clinicopathological, immunohistochemical, and molecular characteristics of uterine MLA and determined their relationships and prognostic significance.
Biomedicines 2023, 11, x FOR PEER REVIEW 4 of 23 variation (CNV) log2 ratios were generated using a depth of coverage normalized that of normal uterine tissues.

PD-L1 22C3 Pharmdx IHC
FFPE tissue blocks were cut into 4 µm sections which were subsequently mounted on Superfrost Plus Microscope Slides (Thermo Fisher) dried at 60 • C for 1 hour.PD-L1 IHC was carried out on a Dako Autostainer Link 48 system (Agilent Technologies) using a Dako PD-L1 IHC 22C3 pharmDx kit (Agilent Technologies), with EnVision FLEX visualization system [46].PD-L1 protein expression was assessed using the combined proportion score (CPS).CPS was calculated as the number of PD-L1-stained cells (tumor cells, lymphocytes, and macrophages) divided by the total number of tumor cells, multiplied by 100.The specimen was considered to have positive PD-L1 expression if CPS ≥ 1 [35,42,46].

IHC Interpretation
PD-L1 immunoreactivity in uterine MLA was assessed using the CPS interpretation guideline, as previously described [47].CPS was calculated as the number of PD-L1stained cells (viable tumor cells, lymphocytes, and macrophages) divided by the total number of viable tumor cells, multiplied by 100.A minimum of 100 viable tumor cells was considered adequate for evaluating PD-L1 positivity.For tumor cells, partial or complete membranous staining at any intensity was regarded as a positive expression.Membranous and/or cytoplasmic staining at any intensity was regarded as positive for tumor-associated immune cells.MMR protein expression in uterine MLA was classified into three categories: preserved, loss, and subclonal loss [41,48,49].A lack of one or more MMR protein expression was defined as MMRd, and preserved expression of all four MMR proteins was defined as MMRp.We regarded the complete absence of nuclear staining (0%), in the tumor cells with appropriate internal control staining (positive nuclear expression in the stromal non-neoplastic cells or lymphocytes), as loss of expression [32].

Statistical Analysis
Pearson's chi-squared test, Fisher's exact test, or linear-by-linear association test was used to determine the association between recurrent pathogenic mutations and clinicopathological characteristics of patients with uterine MLA.Univariate survival analysis, with the log-rank test and Kaplan-Meier plots (RFS and OS), was conducted to evaluate the prognostic significance of recurrent pathogenic mutations and clinicopathological characteristics.Multivariate survival analysis was performed using the Cox proportional hazards model (95% confidence interval) with the backward stepwise elimination method.All statistical analyses were performed using IBM SPSS Statistics for Windows, version 23.0 (IBM Corporation, Armonk, NY, USA).Statistical significance was defined as p < 0.05.

PD-L1 Expression, MMR Protein Expression, MSI Status, and Tumor Mutational Burden
Table 3 summarizes the results of PD-L1 expression by IHC and MMR protein detection by multiplex PCR for MSI status determination.Representative photomicrographs, illustrating immunostaining, are depicted in Figure 4.In the majority of cases (14/17), PD-

PD-L1 Expression, MMR Protein Expression, MSI Status, and Tumor Mutational Burden
Table 3 summarizes the results of PD-L1 expression by IHC and MMR protein detection by multiplex PCR for MSI status determination.Representative photomicrographs, illustrating immunostaining, are depicted in Figure 4.In the majority of cases (14/17), PD-L1 in tumor and tumor-associated inflammatory cells (CPS 0) was not expressed.In the three cases that did express PD-L1, the PD-L1 CPS values were 0.5 (2/3) and 0.1 (1/3).In one of the cases with PD-L1 CPS 0.5, some tumor and inflammatory cells expressed PD-L1 with variable staining intensity.On high-power view, the cells exhibited weak membranous PD-L1 immunoreactivity as well as a paranuclear dot-like and Golgi staining pattern.In the other case with CPS 0.5, stromal lymphocytes and plasma cells expressed PD-L1.While the neoplastic glands were negative for PD-L1, the inflammatory cells surrounding the neoplastic glands and clusters of tumor cells reacted with PD-L1.In the case with PD-L1 CPS 0.1, only a small number of tumor cells expressed PD-L1 with weak-to-moderate staining intensity.Regarding MMR protein expression, all cases (17/17) retained MMR protein staining (for all four MMR proteins), indicating MMRp.Consistent with this finding, 15 cases tested for MSI were interpreted as MSS.Tumor mutational burden was measured using NGS (Table 3).We observed low tumor mutational burden in all examined cases, with a megabase range of 2.70 to 4.72 (median = 3.57).

Clinicopathological and Prognostic Significance of Recurrent Pathogenic Mutations
Based on the pathogenic mutational status, there were no significant differences in clinicopathological characteristics (Table 4) and patient outcomes (Table 5).Univariate survival analysis revealed that initial serosal extension (p < 0.001) and initial or recurrent lung metastasis (p = 0.002) were significant predictors for RFS.Using multivariate survival analysis, we found that serosal extension and initial or recurrent lung metastasis were independent factors for RFS prediction, with hazard ratios of 6.30 (p = 0.037) and 7.31 (p = 0.02), respectively.None of the clinicopathological characteristics and recurrent pathogenic mutations were significantly associated with OS.Figures 5 and 6 display Kaplan-Meier plots, with RFS and OS stratified by clinicopathological characteristics and pathogenic mutations, respectively.Abbreviations: KRAS-Kirsten rat sarcoma viral oncogene homolog; LN-lymph node; PIK3CA-phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha; PTEN-phosphatase and tensin homolog deleted on chromosome 10.

Discussions
In this study, we investigated the genetic features of 17 uterine MLA cases from a single institution.Seventeen primary and four matched metastatic tumors were available for SNV analysis.The most frequently mutated gene was KRAS, followed by PTEN, PIK3CA, GNAQ, CTNNB1, and ARID1A.The obtained results were consistent with that of previous data [3,40,[50][51][52], where KRAS mutations were found in the majority of uterine MLA cases (16/21).In recent studies, we demonstrated that all examined uterine (6/6) and ovarian (4/4) MLA cases harbored pathogenic KRAS mutations [3,10].In addition, we summarized previously reported genetic abnormalities associated with ovarian MLA [10] and found that the most frequently mutated gene was KRAS (23/28), while PIK3CA, NRAS, and ARID1A mutations were uncommon.da Silva et al. [52] reported that the majority of the ovarian and uterine MLAs harbored mutations in KRAS (25/28) and genes frequently mutated in Mullerian tumors, including PIK3CA, CTNNB1, and PTEN.Collectively, the data support the concept found by Kolin et al. [51], which states that KRAS mutations can be considered as one of the defining features of MLA.Although KRAS mutations are not unique to MLA [3,10,39,40,53] and can be identified in 20-30% of the endometrial endometrioid carcinoma [54][55][56], with the additional use of histological features and immunophenotypes characteristic of MLA, the identification of KRAS mutations can confirm MLA diagnosis.
We demonstrated that both the primary and matched metastatic tumors harbored identical KRAS (3/4) and PTEN (1/4) mutations, and did not harbor any additional mutations.da Silva et al. [52] analyzed the mutational profiles of two primary uterine MLAs and their respective distant metastases.Consistent with our result, in one case, both the primary and metastatic tumors harbored an identical KRAS mutation, without additional mutations.However, in the other case, the metastatic tumor exhibited identical KRAS mutations to the primary tumor and harbored additional mutations in genes related to the mitogen-activated protein kinase pathway (mitogen-activated protein kinase kinase kinase 13, mesenchymal epithelial transition, and mitogen-activated protein kinase 3).The latter finding suggests that the progression from primary to metastatic MLA may involve the acquisition of additional mutations.
To the best of our knowledge, this study is the second to report PD-L1 expression in uterine MLA.Horn et al. [59] first reported PD-L1 negativity in four cases of uterine MLA.In line with this finding, we observed that the majority of cases (14/17) exhibited a complete lack of PD-L1 immunoreactivity in both the tumor and immune cells (CPS 0).The remaining three cases focally expressed PD-L1, with CPS < 1.In addition, our observations of retained MMR protein expression and MSS in all examined cases are consistent with our previous findings [41].We re-examined some uterine MLA cases misinterpreted as MMRd during the initial diagnosis and found that they should have been interpreted as MMRp tumors, confirming that they were MSS.A PD-1 inhibitor, pembrolizumab, was approved for advanced, recurrent, or metastatic MSI-H/MMRd ECs as a second-line treatment.Recently, the combination of pembrolizumab with an oral multikinase inhibitor, lenvatinib, has shown remarkable results, with an objective response rate of 36% and median OS of 16.4 months for advanced non-MSI-H/MMRp ECs [24].Although the effectiveness of lenvatinib and pembrolizumab combination therapy for uterine MLA has not been fully elucidated, there have been a few case reports documenting excellent and durable responses to this combination therapy in patients with uterine MLA [60][61][62].Considering that the treatment of patients with advanced or recurrent uterine MLA may differ depending on MSI status, MMR IHC in uterine MLA requires careful interpretation.Repeat IHC and MSI testing may improve diagnosis of challenging cases.
The Proactive Molecular Risk Classifier for Endometrial Cancer introduced four subgroups of EC: (1) DNA polymerase epsilon, catalytic subunit (POLE)-mutant subgroup, harboring mutations in the exonuclease domain in exons 9-14; (2) MMRd subgroup, showing the loss of expression for one or more MMR proteins; (3) p53-abnormal subgroup, demonstrating aberrant p53 expression pattern indicating pathogenic TP53 mutation; and (4) no specific molecular profile (NSMP) subgroup [63][64][65].Even though the vast majority of NSMP ECs is low-grade endometrioid carcinoma, the NSMP subgroup also encompasses high-grade EC, clear cell carcinoma, undifferentiated carcinoma, carcinosarcoma, and MLA [36,40,50,51,53,57,[65][66][67][68].In this study, the majority of MLA cases harbored activating KRAS mutations but not POLE-mutant signatures, MMR deficiency, or TP53 mutation; confirming that these cases belong to the NSMP subgroup.Since the biological behavior of uterine MLA is consistently described as aggressive [36,67,69], the inclusion of this entity in the high-risk non-endometrioid group appears to be justified.Several studies have identified that the high expression of the L1 cell adhesion molecule (L1CAM) and CTNNB1 mutation can add significant prognostic information to the molecular classification of EC [70][71][72][73][74].The significance of L1CAM expression or CTNNB1 mutation in MLA patients has not been elucidated.Based on relatively poor MLA prognosis, compared to other histological types belonging to the NSMP subgroup [3,67], it is likely to exhibit L1CAM overexpression or harbor the CTNNB1 mutation.Further studies with a larger cohort of uterine MLA patients are necessary to clarify the clinicopathological and prognostic significance of L1CAM expression and the CTNNB1 mutation.
We found that 2 of the 17 uterine MLA cases harbor frameshift TP53 mutations.One patient with stage IA MLA did not experience recurrence and is alive at 63.5 months postoperatively, while the other patient with stage IA disease developed recurrence at 8.0 months after surgery and died of disease at 25.1 months postoperatively.The TP53 mutation is known to be extremely uncommon in malignant mesonephric lesions [37,52,57,75] and its clinicopathological significance in patients with uterine MLA has not yet been investigated.We recently experienced a case of dedifferentiated uterine MLA harboring the TP53 mutation [39].IHC revealed that the dedifferentiated component overexpressed p53, while the MLA component exhibited a wild-type p53 expression pattern.We cannot conclude the prognostic implication of the TP53 mutation in uterine MLA because only a few cases were found.However, based on our observations that the majority of uterine MLA cases did not harbor the TP53 mutation and that the clinical course of the two patients with early-stage, TP53-mutant MLA was inconsistent, we believe that there is an urgent need to determine the clinicopathological and prognostic significance of the TP53 mutation in uterine MLA.Similar to previous results showing that the TP53 mutation in 'multiple-classifier' EC cases does not significantly affect disease development or prognosis [76], we hypothesize that the TP53 mutation occurs during late stage of MLA progression and does not affect the molecular landscape, since only two frameshift TP53 mutations were detected in this study of uterine MLAs with pathogenic KRAS mutations and the two cases displayed different outcomes.Further investigations are required to investigate whether TP53 mutation as a significant predictor for patient outcome of uterine MLA or merely a passenger event with no impact on biological behavior.
Therapeutic strategies tailored to both the genetic and epigenetic features of EC are the basis of precision medicine in gynecological oncology [77].Aberrant expression of several cancer-related gene sets has been consistently reported to be a significant contributor for EC progression [78].In addition to those mutational changes, epigenetic alterations, including methylation, acetylation, and phosphorylation of nuclear chromatin, play a central role in EC development and progression [78,79].Particularly, non-coding RNAs (ncRNAs) are involved in the regulation of cellular metabolism, growth, and neoplastic transformation [78,80].They have very little or no protein-coding capability [81,82], but their expression patterns can modulate the function of oncogenes and tumor suppressors, resulting in either promotion or suppression of tumorigenesis and progression [77].Their regulation of gene expression can occur in different steps, at epigenetic, transcriptional, and post-transcriptional levels [78].Accumulating evidence shows that the abnormal expression of ncRNAs is associated with the prevalence and prognosis of many different types of human cancers [82,83].Some deregulated ncRNAs have recently been suggested as potential risk factors that can better define the biological behavior of EC and be used as prognostic markers to guide the risk stratification of EC patients.It is surprising to note that the association between ncRNA and EC has only recently been emerging in the literature, and that most of the papers regarding this association have been concentrated into the last three years [77].The following competing endogenous RNAs seemed to be associated with poor prognosis of EC: AC074212.1,ADARB2-AS1, C2orf48, C8orf49, C10orf91, FER1L4, FP671120.4,GLIS3-AS1, HOXB-AS1, LINC00483, LINC00491, LINC01143, LINC01352, LINC01410, LINC02381, MIR503HG, PCAT1, RP11-357H14.17and RP11 89K21.1 [77].In contrast, LINC00237, LINC00475, LINC00958, and LNCTAM34A were reported to exhibit a favorable prognostic effect in EC patients.The expressions of these molecules were found to be deregulated in EC compared to normal endometrial tissue.Identifying deregulated microRNAs (miRs) remains an ongoing endeavor.miRs are short molecules of ncRNA that function as post-transcriptional regulators of gene expression [84].In a meta-analysis by Delangle et al. [85], a number of significantly deregulated miRs were identified and classified as onco-miRs, suppressor miRs, and those with discordant functions.Similarly, a recent systematic review by Bloomfield et al. [86] revealed the deregulated levels of circulating miRs in the serum and plasma of EC patients.These studies suggest that adequate combinations of miR expression with conventional pathological parameters of EC may serve as prognostic markers that can help in predicting risk stratification of patients.Taken together, epigenetic modifications are gaining increasing importance for the characterization of EC.A group of molecules is emerging as identifiable risk factors that aid in establishing an accurate diagnosis and assessing clinical prognosis.Particularly, there is a significant correlation between the alterations of several ncRNAs and miRs and the clinical course of EC patients, representing the possibility of including these molecules in stratifying patients at greater risk of relapse and worse outcome [82].A comprehensive analysis of these molecules is the way to pursue towards personalized medicine, in which each patient is characterized by a specific set of epigenetic alterations, whose targets are well defined, and for whom drawing a therapeutic strategy would yield better results [80,82].

Conclusions
We comprehensively investigated the clinicopathological and immunophenotypical features of 17 consecutive cases of uterine MLA from a single institution.We confirmed that uterine MLA is an aggressive malignancy, showing advanced stage, frequent postoperative disease recurrence, and frequent lung metastasis.Initial serosal extension and lung metastasis were independent prognostic factors for RFS prediction, while none of the clinicopathological or molecular features were significantly associated with OS of uterine MLA patients.IHC revealed that none of the cases overexpressed PD-L1 or were MMR deficient.We conducted targeted sequencing to analyze the molecular features of uterine MLA.We found that the majority of cases harbored pathogenic KRAS mutations.Two cases harboring the frameshift TP53 mutation were also identified, but the clinicopathological and prognostic significance of TP53 mutation could not be determined.The most frequent abnormalities were gains of chromosome 1q, 2, 10, and 20.Both clinicians and pathologists should be aware of these features to establish an accurate diagnosis of uterine MLA and to ensure appropriate therapeutic management of this rare entity.

Informed Consent Statement:
Regarding the retrospective nature of this study, the Institutional Review Board waived the requirement for the investigators to obtain signed informed consent.

Figure 1 .
Figure 1.Histological and immunophenotypical features of uterine mesonephric-like adenocarcinoma.(A) Tubular and glandular patterns showing compactly aggregated small-to-medium-sized tubules and elongated ductal structures.(B) Solid and tubular patterns showing solid cellular sheets and slit-like tubular lumina.(C) Complex small tubular proliferation with back-to-back arrangement.(D) Eosinophilic, hyaline-or colloid-like intraluminal secretions.(E) Lack of estrogen receptor expression.(F) Wild-type p53 immunostaining pattern.(G) Non-diffuse p16 positivity.(H) Moderate-to-strong nuclear immunoreactivity for transcription termination factor 1. (I) Uniform and strong GATA-binding protein 3 expression.

Figure 1 .
Figure 1.Histological and immunophenotypical features of uterine mesonephric-like adenocarcinoma.(A) Tubular and glandular patterns showing compactly aggregated small-to-medium-sized tubules and elongated ductal structures.(B) Solid and tubular patterns showing solid cellular sheets and slit-like tubular lumina.(C) Complex small tubular proliferation with back-to-back arrangement.(D) Eosinophilic, hyaline-or colloid-like intraluminal secretions.(E) Lack of estrogen receptor expression.(F) Wild-type p53 immunostaining pattern.(G) Non-diffuse p16 positivity.(H) Moderateto-strong nuclear immunoreactivity for transcription termination factor 1. (I) Uniform and strong GATA-binding protein 3 expression.

Figure 3 .
Figure 3. Diagram depicting chromosomal copy number variations, determined using array comparative genomic hybridization.Copy number variations are indicated in green or red for gain or loss in copy number, respectively.All cases of uterine mesonephric-like adenocarcinoma exhibit gains of chromosome 1q (yellow arrow) and 10 (orange arrow), and most of the cases exhibit gains of chromosome 2 (blue arrow) and 20 (purple arrow).

Figure 3 .
Figure 3. Diagram depicting chromosomal copy number variations, determined using array comparative genomic hybridization.Copy number variations are indicated in green or red for gain or loss in copy number, respectively.All cases of uterine mesonephric-like adenocarcinoma exhibit gains of chromosome 1q (yellow arrow) and 10 (orange arrow), and most of the cases exhibit gains of chromosome 2 (blue arrow) and 20 (purple arrow).

Figure 5 . 23 Figure 6 .
Figure 5. Kaplan-Meier plots showing probability of recurrence-free survival (RFS) stratified by clinicopathological characteristics and pathogenic mutations.Initial serosal extension and lung

Figure 6 .
Figure 6.Kaplan-Meier plots showing probability of overall survival (OS) stratified by clinicopathological characteristics and pathogenic mutations.None of the examined parameters significantly predict worse OS of patients with uterine MLA.KRAS-Kirsten rat sarcoma viral oncogene homolog; PIK3CA-phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha; PTENphosphatase and tensin homolog deleted on chromosome 10.