Somatic mutation analysis of Mesonephric-Like adenocarcinoma and associated putative precursor Lesions: Insight into pathogenesis and potential molecular treatment targets

Elizabeth Arslanian: Investigation, Writing – original draft. Kamaljeet Singh: Writing – review & editing. C. James Sung: Writing – review & editing. M. Ruhul Quddus: Conceptualization, Methodology, Writing – review & editing, Supervision.

The expenses were covered by the departmental education fund; no outside funding was obtained.

We retrospectively identified cases by searching the keyword “mesonephric” in pathology reports from January 2015 to July 2021 in our institution’s database. KS and MRQ reviewed haematoxylin & eosin (HE) and immunohistochemistry (IHC) slides. We included in-house tumors fulfilling diagnostic criteria for MLA as per the 2020 World Health Organization (WHO) Classification of Female Genital Tumours (n = 3) ( WHO Classification of Tumours Editorial Board, 2020 ). The morphologic findings of two of these cases were reported in 2021 ( Kulkarni et al., 2021 ). One case was reported as a mesonephric carcinoma. However, upon review, we believe it is more consistent with a diagnosis of MLA. We obtained Institutional Review Board approval. Clinical information was retrieved from the electronic medical record. Immunohistochemistry studies were performed as part of the diagnostic workup ( Table 1 ). We used the following antibody panel in all samples: GATA3, TTF-1, estrogen receptor (ER), CD10 and p16 INK4a antigen. Other markers were evaluated in only one or two of the cases. Table 1 Immunostaining results. IHC stain Clone and vendor Case 1 Case 2 Case 3 GATA3 L50-823, Cell Marque Positive Positive Rare positive foci TTF-1 8G7G3/1, Dako Negative Negative Positive ER EP1, Dako Negative Negative Focally positive PR PgR 1294, Dako Negative N/A N/A CD 10 56C6, Dako Focal apical membrane staining Focal patchy positive Negative p16 BC42, BioCare Medical Focal staining Focal patchy positive Patchy positive PAX-8 BC12, Biocare Medical N/A N/A Strong and diffuse Calretinin DAK-Calret 1, Dako N/A N/A Negative p53 DO-7, Dako Wild-type pattern N/A Wild-type pattern Ki-67 MIB-1, Dako Focally increased N/A > 60 % Synaptophysin DAK-SYNAP, Dako Negative N/A Negative Chromogranin A DAK-A3, Dako Negative N/A N/A CD56 123C3, Roche N/A N/A Focally positive Cytokeratin (pan) AE1& AE3, Dako N/A N/A Patchy positive Cytokeratin CAM 5.2 CAM5.2, BD Biosciences Positive N/A Patchy positive Cytokeratin 7 OV-TL12/30, Dako N/A N/A Strong and diffuse Cytokeratin 20 Ks20.8, Dako N/A N/A Rare positive foci Cytokeratin 5/6 D5/16 B4, Dako N/A N/A Negative EMA E29, Dako N/A N/A Patchy positive CA 125 M11, Dako N/A N/A Positive hMLH1 ES05, Dako Intact N/A Intact hMSH2 FE11, Dako Intact N/A Intact hMSH6 EP49, Dako Intact N/A Intact PMS2 EP51, Dako Intact N/A Intact CDX2 DAK-CDX-2, Dako N/A N/A Negative Inhibin alpha R1, Dako N/A N/A Negative WT1 6F-H2, Dako N/A N/A Negative Alpha fetoprotein Polyclonal, Cell Marque N/A N/A Negative AR SP107, Cell Marque N/A N/A Negative Desmin D33, Dako N/A N/A Negative Vimentin V9, Dako N/A N/A Positive N/A: not performed. Immunostaining results. N/A: not performed. We chose one to three block(s) of interest for every case, each either displaying MLA or an associated finding, namely endometriosis and adenofibroma. When a tissue block with a pure concomitant lesion (without MLA being present on the same block) was not available, we selected a block representing both the associated lesion and the MLA. Each sample consisted of fifteen unstained sections and one HE slides from the same formalin-fixed paraffin embedded (FFPE) tissue block. The unstained 10 µm-thick sections were prepared on uncharged slides without additives under DNA extraction precautions. Microdissection for tumour enrichment was not conducted. Next generation sequencing using Centogene US LLC’s solid tumour panel was performed. This panel covers 149 genes (full sequencing of 106 genes and hotspot analysis of 43 genes), with a > 97 % > 200x coverage. Targeted genes are the following: ABL1, AKT1, AKT2, AKT3, ALK, APC, AR, ARAF, ARID1A, ASXL1, ATM, ATR, ATRX, AXL, BAP1, BRAF, BRCA1, BRCA2, BTK, CBL, CCND1, CDH1, CDK12, CDK4, CDK6, CDKN1B, CDKN2A, CDKN2B, CHEK1, CHEK2, CREBBP, CSF1R, CTNNB1, DDR2, EGFR, ERBB2, ERBB3, ERBB4, ERCC2, ESR1, EZH2, FANCA, FANCD2, FANCI, FBXW7, FGFR1, FGFR2, FGFR3, FGFR4, FLT3, FOXL2, GATA2, GNA11, GNAQ, GNAS, H3-3A, H3C2, HNF1A, HRAS, IDH1, IDH2, JAK1, JAK2, JAK3, KDR, KEAP1, KIT, KMT2A, KMT2C, KMT2D, KNSTRN, KRAS, MAGOH, MAP2K1, MAP2K2, MAP2K4, MAPK1, MAX, MDM4, MED12, MEN1, MET, MLH1, MPL, MRE11, MSH2, MSH6, MTOR, MYC, MYCN, MYD88, NBN, NF1, NF2, NFE2L2, NOTCH1, NOTCH2, NOTCH3, NRAS, NTRK1, NTRK2, NTRK3, PALB2, PDGFRA, PDGFRB, PIK3CA, PIK3CB, PIK3R1, PMS2, POLE, PPP2R1A, PTCH1, PTEN, PTPN11, RAC1, RAD50, RAD51, RAD51B, RAD51C, RAD51D, RAF1, RB1, RBM10, RET, RHEB, RHOA, RIT1, RNF43, ROS1, SETD2, SF3B1, SLX4, SMAD4, SMARCA4, SMARCB1, SMO, SPOP, SRC, STAT3, STK11, TERT, TOP1, TP53, TSC1, TSC2, TSHR, U2AF1, VHL, XPO1 .

This is an academic institution and all patients receiving medical care here sign a consent form agreeing that their material can be used for educational purposes without using any unique identifiers.

Three cases of MLA were included, arising in the ovary (n = 2) and in the uterine corpus (n = 1). Clinical findings are summarized in Table 2 . Patient age at diagnosis ranged from 65 to 67 years, with a median of 66 years. Median follow-up was 32 months. Patient 3 underwent cancer genetic testing which was negative for pathogenic mutations; a variant of uncertain significance in the ATM gene was found. Table 2 Clinical findings. Case Presentation Site Concurrent lesions FIGO Stage Treatment Follow-up 1 66 y.o., 8-cm pelvic mass Right ovary Endometriotic cyst, right ovary. Endometrial atypical hyperplasia. IIIA1 TAH-BSO, infracolic omentectomy and right para -aortic lymph node biopsy, adjuvant CTx (Taxol/Carboplatin/Avastin, maintenance Avastin followed by Carbo/Doxil) Deceased 32 months post-surgery 2 65 y.o., 8.5-cm pelvic mass Uterine corpus Endometriosis, bilateral fallopian tubes. IIA2 Radical hysterectomy-BSO, bilateral pelvic lymphadenectomy, adjuvant radiation therapy and CTx (Cisplatin) No recurrence 33 months post-surgery 3 67 y.o., 18-cm pelvic mass and postmenopausal bleeding Left ovary Endometriosis and seromucinous cystadenofibroma with focal borderline changes, left ovary. IC TAH-BSO, bilateral pelvic node dissection and omentectomy, adjuvant CTx No recurrence 15 months post-surgery TAH indicates total abdominal hysterectomy; BSO, bilateral salpingo-oophorectomy; CTx: chemotherapy; Y.o.: years old. Clinical findings. TAH indicates total abdominal hysterectomy; BSO, bilateral salpingo-oophorectomy; CTx: chemotherapy; Y.o.: years old.

We report the molecular alterations of three MLAs and associated lesions. We believe our most interesting finding is the identification of a potentially actionable target, the PIK3CA H1047Q variant. Mutations in the PIK3CA gene have been formerly described in KRAS -mutated MLA. To the best of our knowledge, this variant has not been previously reported. Other variants in the same hotspot predict response to a treatment combination of the ER antagonist fulvestrant and the α-specific PI3K inhibitor alpelisib in ER-positive advanced breast carcinoma. This medication combination was shown to extend progression-free and overall survival in PIK3CA -mutated cases ( Juric et al., 2019 , André et al., 2021 ). This raises the possibility that variants in H1047 might be predictive biomarkers in carcinomas of other sites; like MLA. This ought to be further investigated, as one might wonder if these drugs could increase life expectancy after an MLA diagnosis. Additionally, the PIK3CA H1047Q variant found in case 3 was also present in the associated cystadenofibroma with focal borderline changes. This supports a relationship between the two lesions. We report KRAS G12D and G12V in MLA. Associated seromucinous cystadenofibroma with focal borderline changes shared the G12V variant. This variant was also detected in the endometriosis sample; however, we could not exclude tumor contamination. According to the 2020 WHO blue book, pathogenesis of seromucinous cystadenofibroma is unknown. KRAS mutations have been described in seromucinous borderline tumors ( WHO Classification of Tumours Editorial Board, 2020 ). Although literature is still limited, KRAS mutations in MLA are well recognized. In a study of 28 MLAs, 89 % of tumors were KRAS -mutated. KRAS mutations are also known to occur in other gynecologic cancers, including cervical mesonephric carcinoma ( da Silva et al., 2021 ). KRAS -mutated endometrial carcinoma histological subtypes include endometrioid, serous, mixed, dedifferentiated and poorly differentiated carcinoma and carcinosarcoma ( Kolin et al., 2019 ). Ovarian tumors harboring KRAS mutations include serous borderline tumor, low-grade serous carcinoma, mucinous cystadenoma/adenofibroma, mucinous borderline tumor, mucinous carcinoma, endometrioid carcinoma, clear cell carcinoma, seromucinous borderline tumor, borderline Brenner tumor and struma ovarii. KRAS mutations are common in MLA, but not specific to this neoplastic process in the gynecologic tract. Their presence in at least some associated Müllerian disease might also support a common pathogenesis. As far as we are aware, there are no guidelines requiring molecular testing for MLA diagnosis ( WHO Classification of Tumours Editorial Board, 2020 ). This entity was introduced a few years following the integrated genomic characterization of endometrial carcinoma by The Cancer Genome Atlas (TCGA) Research Network. This study focused on endometrioid and serous histological subtypes ( Levine, 2013 ). Hence, the ProMisE classifier does not apply to MLA. This algorithm includes evaluation of POLE mutational status, microsatellite stability and p53 status. As discussed by Pors et al., MLA is usually microsatellite stable with a wild type p53 status ( Pors et al., 2021 , Levine, 2013 ). As far as we are aware, POLE mutations have not been reported in MLA either. This tumor would fall within the “copy-number low” molecular category. Our findings also support the currently limited evidence of MLA’s Müllerian differentiation ( WHO Classification of Tumours Editorial Board, 2020 , da Silva et al., 2021 , Pors et al., 2021 , Chapel et al., 2018 Sep , McCluggage et al., 2020 Jan , Dundr et al., 2020 ). This distinguishes MLA from mesonephric carcinoma. The latter is thought to arise from mesonephric remnants ( WHO Classification of Tumours Editorial Board, 2020 ). Proof of Müllerian origin in our cases includes concurrent Müllerian neoplasia and endometriosis. This association is already known. Stage and follow-up information was presented in result Table 2 . Case 1 (stage FIGO IIIA1; deceased 32 months postoperatively) provides further evidence of MLA’s possible dismal prognosis. Further follow-up is required in cases 2 and 3 to determine their long-term disease-related outcome. Overall, we provide further evidence of KRAS mutations in MLA, sometimes concurrent with PICK3CA mutations. One of our MLA cases shares a PIK3CA variant in the potentially actionable H1047 hotspot with a seromucinous cystadenofibroma with focal borderline change. To the best of our knowledge, this is the first description of an alteration in this hotspot in MLA. As previously reported in the literature, coexisting Müllerian neoplasia, sometimes with proof of shared molecular origin, support a Müllerian rather than a mesonephric differentiation.

Gross evaluation revealed an 8-cm solid and cystic unilateral right ovarian mass, adhering to the uterine serosa and invading the subserosal myometrium. Histologic examination of the right ovary confirmed the presence of tubular and solid MLA with focal necrosis. Nuclei were crowded, with open chromatin. Immunostains supported the diagnosis ( Fig. 1 , Table 1 ). There was also an endometriotic cyst involved by carcinoma in the right ovary. Endometrial atypical hyperplasia was identified. The cervix, left ovary, bilateral fallopian tubes and omentum appeared benign. A right paraaortic lymph node biopsy was positive for metastatic MLA. No other nodes were sampled. Fig. 1 Case 1. A. Mesonephric-like adenocarcinoma (MLA) in tubular and solid patterns with focal necrosis. B. CD10, focal apical membranous staining. C. GATA3, positive. D. ER, negative. Case 1. A. Mesonephric-like adenocarcinoma (MLA) in tubular and solid patterns with focal necrosis. B. CD10, focal apical membranous staining. C. GATA3, positive. D. ER, negative. The G12V (c.35G > T p.(Gly12Val)) pathogenic KRAS variant was identified in MLA ( Table 3 ). Endometriosis was analyzed separately. It exhibited the KRAS G12V variant. However, a piece of tumor noted in the sample could account for the mutation. Table 3 Molecular findings. Case Lesion KRAS variant PIK3CA mutation 1 Mesonephric-like carcinoma G12V N.D. 1 Endometriotic cyst G12V * N.D. 2 Mesonephric-like carcinoma G12D N.D. 3 Mesonephric-like carcinoma G12V H1047 hotspot 3 Seromucinous cystadenofibroma with focal borderline changes G12V H1047 hotspot N.D.: not detected. * Contamination by mesonephric-like carcinoma favoured/not excluded. Molecular findings. N.D.: not detected. Contamination by mesonephric-like carcinoma favoured/not excluded. Macroscopic findings included the presence of a necrotic and haemorrhagic 8.5-cm myometrial mass. Microscopic examination confirmed the diagnosis of MLA ( Fig. 2 , Table 1 ) based in the uterine corpus and extending into the lower uterine segment and cervix. The lesion grew in a glandular pattern, with necrosis and intraluminal colloid-like secretions. Tumor nuclei were crowded, with dense chromatin. The entire cervix was submitted, and mesonephric remnants were not identified. Focal mucosal endometriosis was noted in both fallopian tubes. Right and left ovaries were unremarkable. Bilateral pelvic lymph node regional resection was negative for metastatic carcinoma. Fig. 2 Case 2. A and B. MLA in glandular pattern with necrosis. C. GATA3, positive. D. ER, negative. Case 2. A and B. MLA in glandular pattern with necrosis. C. GATA3, positive. D. ER, negative. MLA exhibited the KRAS G12D variant (c.35G > A p.(Gly12Asp)). The specimen consisted of an 18-cm disrupted solid and cystic unilateral left ovarian mass. Cystic cavities contained clear yellow mucinous fluid and brown serous fluid. No necrosis nor haemorrhage was identified. The tumor exhibited solid and tubular histological patterns, with intraluminal eosinophilic colloid-like material. Nuclei were crowded, with dense chromatin. Morphologic and immunohistochemical findings supported the diagnosis of MLA ( Fig. 3 , Table 1 ). Endometriosis and a seromucinous cystadenofibroma with focal borderline changes were also present in the ipsilateral ovary. Fig. 3 Case 3. A and B. MLA in solid and tubular patterns C. Seromucinous cystadenofibroma with focal borderline changes. D. TTF-1, positive. E. GATA3, rare positive foci. Case 3. A and B. MLA in solid and tubular patterns C. Seromucinous cystadenofibroma with focal borderline changes. D. TTF-1, positive. E. GATA3, rare positive foci. Molecular testing revealed a pathogenic KRAS G12V variant (c.35G > T p.(Gly12Val)), with a relatively high allele frequency (86.3 % out of 1260 NGS reads). Besides the KRAS alteration, a PIK3CA mutation in the H1047 hotspot (c.3141 T > A p.(His1047Gln) variant) was detected, both in the MLA and in the cystadenofibroma with focal borderline changes. Both lesions were analyzed separately.

Mesonephric-like adenocarcinoma (MLA) is a recently described malignancy of the gynecologic tract. MLA is a rare adenocarcinoma subtype affecting mostly postmenopausal patients. It has been reported in the ovary, uterine corpus, vagina and para -adnexal soft tissue. MLA can arise from endometriosis ( WHO Classification of Tumours Editorial Board, 2020 , da Silva et al., 2021 , Pors et al., 2021 ). The association with endometriosis and other Müllerian benign, borderline, and malignant lesions supports a Müllerian origin ( WHO Classification of Tumours Editorial Board, 2020 , da Silva et al., 2021 , Pors et al., 2021 , Chapel et al., 2018 Sep , McCluggage et al., 2020 Jan , Dundr et al., 2020 ). KRAS mutations underpin MLA. Other molecular changes reported in MLA include NRAS, BRAF, PIK3CA, PTEN and CTNNB1 mutations and some copy number variations ( da Silva et al., 2021 , Pors et al., 2021 ). Recent studies have shown that MLA shares molecular alterations with concurrent benign and proliferative Mullerian lesions, suggesting a putative precursor role. To our knowledge, no association with mesonephric remnants has been described to date. It has, however, been hypothesized that some cases develop from paraovarian remnants ( WHO Classification of Tumours Editorial Board, 2020 ). MLA presents with vaginal bleeding or as a solid and/or cystic mass ( WHO Classification of Tumours Editorial Board, 2020 , Pors et al., 2021 ). It can follow an aggressive course ( Pors et al., 2021 , Deolet et al., 2022 ). As the name suggests, MLA’s histologic and immunohistochemical features overlap with those of cervical HPV-independent mesonephric type adenocarcinoma. Solid, papillary, tubular and glandular patterns are seen in MLA. Intraluminal amorphous eosinophilic material is characteristic. MLA is usually positive for GATA3 and/or TTF-1, and negative for hormone receptors. Estrogen receptor (ER) can be focally expressed in some cases. Apical membranous CD10 expression can be seen ( WHO Classification of Tumours Editorial Board, 2020 , Pors et al., 2021 ). Some tumors show both GATA3-positive, TTF-1 negative and TTF-1 positive, GATA-3 negative areas. Cervical mesonephric carcinoma can also express these markers. TTF-1 seems to be positive in a lower proportion of mesonephric carcinomas than in MLAs. The peculiar inverse pattern of GATA-3 and TTF-1 positivity does not appear to be found in mesonephric carcinoma ( Pors et al., 2018 ). MLA, however, differs from mesonephric carcinoma by its location and absence of associated precursor mesonephric remnants or hyperplasia. Mesonephric remnants may be found in the cervical lateral walls. Some authors have also suggested their presence in the myometrium, vagina, mesosalpinx and ovarian hilum ( Howitt and Nucci, 2018 Feb ). Others question that they could occur in the uterine corpus wall ( Deolet et al., 2022 ). Our goal was to further define the molecular signature of MLA and associated findings. We hypothesized that MLA and concurrent lesions might have a similar mutational profile. We performed genetic testing on three MLAs and concomitant lesions.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.