Mesonephric-like adenocarcinomas (MLA) are a recently described morphologic subtype of malignancies of the female genital tract sharing histopathologic, immunohistochemical and molecular characteristics of mesonephric carcinomas [Euscher et al. 2020, Mills et al. 20222, Kim et al. 2022, Howitt & Nucci 2018, McCluggage 2022]. Unlike mesonephric carcinomas, they are not associated with mesonephric remnants [Euscher et al. 2020, Mills et al. 20222, Kim et al. 2022, Howitt & Nucci 2018, McCluggage 2022]. Histopathologically, they display mixed morphologic features on H&E-staining (Fig. 1a-d [Mirkovic et al. 2018, Pors et al. 2018, Horn et al. 2020, Euscher et al. 2020]). On immunohistochemistry, MLA show negativity or limited positivity for estrogen receptor staining and positive staining for TTF-1, CD 10 (Fig. 1e-g) and GATA-3 in most cases as well as for calretinin in some cases [Mirkovic et al. 2018, Euscher et al. 2020, Park et al. 2022, McCluggage 2022].
Adenocarcinomas with mesonephric-like features mostly occur within the uterus (74.7%; 115/154), a minority are diagnosed within the ovary (25.3%; 39/154 [Deolet et al. 2021]). A similar distribution was reported in a multi-intitutional study [Pors et al. 2021]: 63.8% (44/69) were of endometrial and 36.2% (25/69) of ovarian origin. Single cases have been described in the fallopian tube [Xie et al. 2021] or mesocolon [Deolet et al. 2022] in association with endometriosis. There are no obvious histomorphologic/immunohistochemical differences between the tumors located in the uterus versus those within the ovaries in the cases reported until now [Mircovic et al. 2018, McCluggage 2022, Deolet et al. 2021, Koh et al. 2022].
Ovarian MLA may be associated with endometriosis or other epithelial ovarian tumors [Seay et al. 2020, McCluggage et al. 2020, Chen et al. 2020, Chang et al. 2023].
Based on immunohistochemical and molecular findings (see below) it is suggested that MLA are Müllerian-derived pathogenetically and display morphologic mesonephric features, indicating a process of transdifferentiation [daSilva et al. 2021, Euscher et al. 2020, McCluggage 2022]. The concept of transdifferentiation may be supported by the presence of non-malignant mesonephric-like proliferations seen in association with mucinous ovarian tumors and identical molecular alterations within the mucinous borderline tumor and associated ovarian MLA [Nilforoushan et al. 2022].
With respect to the molecular classification of endometrial carcinomas [Soslow et al. 2019], uterine MLA show p53 wildtype immunostaining, retained mismatch-repair protein expression (see Fig. 1h-j) and are not assciated with a POLE-mutation [Kolin et al. 2019, Horn et al. 2020, Ma et al. 2022, Deolet et al. 2021] defining them as endometrial carcinomas with no special molecular profile (NSMP [Horn et al. 2020, Deolet et al. 2021, McCluggage 2022]). Within ovarian MLA they show a retained mismatch-repair protein expression [Koh et al. 2022] and p53 wildtype staining [Koh et al. 2022, Chen et al. 2020, Mircovic et al. 2018].
Uterine MLA showed a significantly reduced progression-free survival in comparison to endometroid endometrial carcinomas, even compared to FIGO high grade (G3) ones (p < 0.001 [Kim et al. 2022]). Patients with endometrial MLA represent a poor clinical outcome, and 60–80% of the patients will be affected by recurrent disease or will die of the tumor [Mirkovic et al. 2018, Euscher et al. 2020, Kim et al. 2022, Kolin et al. 2019, Pors et al. 2021]. Compared to endometrioid and serous endometrial carcinomas, MLA showed the lowest median progression-free survival (183 versus 67.1 versus 18.2 months; p < 0.0001 [Euscher et al. 2020]). Pulmonary involvement, detected in up to 60% of endometrial MLA, is the most common site of distant spread [Kim et al. 2021, Park et al. 2022, da Silva et al. 2021, Euscher et al. 2020]. Hepatic metastatic spread is ten times higher in patients with endometrial MLA compared to endometrioid endometrial carcinomas and five times more frequent when compared to those with serous tumors [Pors et al. 2021, Euscher et al. 2020, Mao et al. 2020]. Other uncommon metastatic sites are the brain, spleen and vertebrae [Pors et al. 2021, Deolet at al. 2021]. Although many cases of endometrial MLA may be associated with adverse prognostic factors, such as large tumor size, deep myometrial involvement, cervical stromal invasion, involvement of lymphatic and/or blood vessels and retroperitoneal lymphatic spread [Na & kim 2019, Park et al. 2022, Kim et al. 2022, Euscher et al. 2020], there is an increasing evidence that the presence of a mesonephric-like histology itself may represent an independent poor prognostic feature [Euscher et al. 2020, Pors et al. 2021, Al Nabhani et al. 2022].
For ovarian MLA, Koh et al. [2022] reported a disease-free survival of 24.5 months (N = 5), and Pors et al. [2021] described a progression-free survival of 68% and an overall survival of 71% in a multi-institutional study of 25 cases. Three patients with FIGO-stage IC2 who were treated with adjuvant chemotherapy (carboplatin plus paclitaxel) showed no evidence of disease (NED): one patient after 13 months [Chen et al. 2020], one after eight [Xie et al. 2021] and one additional patient after 13 months [Xie et al. 2021] of follow-up, respectively. Within the study of Deolet et al. [2022], including patients with different FIGO-stages, 3/5 showed NED after a mean follow-up of 22.4 months (range 8–46 months), 1/5 recurred after 11 months, and one patient presented with a partial response in FIGO-stage IVB after polychemotherapeutic treatment at eight months follow-up. That very limited data for ovarian MLA may suggest a more favorable prognosis when compared to uterine MLA, but further investigations are necessary addressing that feature.
Overall, the most frequent mutational event in MLA of the female genital tract is a KRAS-alteration [Mircovic et al. 2018, Kolin et al. 2019, Na & Kim 2019, Horn et al. 2020, Ma et al. 2022, da Silva et al. 2021, Koh et al. 2022, Deolet et al. 2022].
25–30% of the endometrial MLA cases harbour additional mutational events within the PTEN-, CTNNB1- and ARID1A-genes [Mirkovic et al. 2018, Kolin et al. 2019, Euscher et al. 2020, da Silva et al. 2021, Mills et al. 2022, Ma et al. 2022] which are also common mutational alterations in endometrioid endometrial carcinomas.
Within the group of non-MLA of the endometrium, about 17% represent KRAS-mutations and are associated with an improved prognosis compared to KRAS-wildtype carcinomas [Kolin et al. 2018]. Because of the high frequency of KRAS-mutational events in MLA (see below [Mirkovic et al. 2018, Kolin et al. 2019, Ma et al. 2022, Koh et al. 2022]), it may be expected that endometrial MLA are associated with a favorable clinical outcome - but the contrary is the case [Euscher et al. 2020, McCluggage 2022].
About 10–25% of ovarian MLA habour a wide range of different additional mutational events beside the KRAS-alteration, including CTNNB1, PTEN, NOTCH3, NRAS and PIK3CA [daSilva et al. 2021, Deolet et al. 2022].
There is a wide range of treatment approaches in patients affected by MLA of the female genitalia [Euscher et al. 2020, Chen et al. 2020, Deolet et al. 2021, Xie et al. 2021, Koh et al. 2022, Deolet et al. 2022].
Within the review of Deolet et al. [2021], the majority of patients with endometrial MLA underwent a total hysterectomy with bilateral salpingo-oophorectomy and pelvic lymph node dissection, in some cases para-aortal lymph nodes were also removed [Mirkovic et al. 2018, Deolet et al. 2021, Horn et al. 2020]. In ovarian MLA, the surgical procedure has rarely been reported. Within the informative cases, tumor debulking was performed including an omentectomy [Deolet et al. 2021, Chen et al. 2020, Koh et al. 2022]. Patients with endometrial MLA and morphologic risk factors received adjuvant radiation. About one fifth of the reprted endometrial as well as ovarian MLA were treated with carboplatin and paclitaxel postoperatively [Deolet et al. 2021, Chen et al. 2020]. As mentioned above both uterine and ovarian MLA are negative or show only limited positivity for steroid hormone receptors (see Fig. 1e [Mircovic et al. 2018, Euscher et al. 2020, Horn et al. 2020, Chen et al. 2020, Kim et al. 2021, Kim et al. 2022, Koh et al. 2022, daSilva et al. 2021, Ma et al. 2022, Park et al. 2022, McCluggage 2022]), suggesting hormonal treatment may not be effective in MLA. One case of endometrial cancer with a mixed endometrioid and MLA histology was treated with progesterone therapy and recurred six years later only with the MLA component [Yano et al. 2019]. Another case, initially (mis-)diagnosed as a low-grade endometrioid type endometrial cancer received hormonal treatment (not further specified) and recured 17 months later in the liver [Euscher et al. 2020]. Reviewing 60 cases, hormonal treatment was not reported for any MLA occuring in the ovary [Koh et al. 2022].
Until now, mesonephric-like histopathology has not been incorporated into clinical guidelines [Euscher et al. 2020, Chen et al. 2020, Kim et al. 2021, Deolet et al. 2021]. Because of its aggressive clinical behavior, close oncologic follow-up is indicated [Euscher et al. 2020, Kim et al. 2022, Pors et al. 2021, Deolet et al. 2021], and chest imaging may be recommended because of the increased frequency of pulmonary spread after histopathological diagnosis of a mesonephric-like phenotype [Mills et al. 2022]. Because of the aggressive behavior, Euscher et al. [2020] suggested that treatment algorithms used for high-grade endometrial carcinomas should be considered in cases with mesonephric-like histopathology, even if they present with low-stage disesase.
Based on the published data, the optimal (neo-)adjuvant systemic treatment remains unknown. Using the morphomolecular approach for the diagnosis of MLA of the female genital tract (mixed morphology on H&E-staining and immunoexpression of mesonephric-like markers; see above, [Horn et al. 2020, Pors et al. 2021, McCluggage et al. 2022]), it has been shown that a KRAS-mutation is a common finding in MLA [Mircovic et al. 2018, Kolin et al. 2019, Horn et al. 2020, Ma et al. 2022, da Silva et al. 2021, Koh et al. 2022, Deolet et al. 2022, McCluggage et al. 2020]. For example, 92% of the uterine MLA (12/13) and 87% of the ovarian MLA (13/15) in the study of da Silva et al. [2021] harbored KRAS somatic mutations affecting the hotspot codons 12 and 13 of KRAS (G12D, G12V, G12C, G12A and G13D).
However, a mutational analysis was either not performed or reported in all published cases of the female genital tract [e.g. Pors et al. 2021, Xie et al. 2021, Chang et al. 2022]. Within the review of Deolet et al. [2021] mutational results were available for 32.5% of the ovarian and uterine MLA included in that study. Reviewing 60 ovarian MLA, KRAS-mutational status was reported for 46.7% (28/60) of the cases [Koh et al. 2022].
The frequency of KRAS-alterations and the distribution of the different mutational sites of the informative published cases of uterine and ovarian MLA are summarized in Fig. 3. After analyzing the first published cases, it was hypothesized that the p.G12yD-alteration may be more prevalent in ovarian MLA while a p.G12V-alteration was more common in uterine tumors [Mirkovic et al. 2018]. However, increasing evidence does not confirm any predilection of any mutational site in the different location of the MLA within the female genitalia (Fig. 3).
There are very limited data on a potential biological significance of the different sites of KRAS-alterations in gynecologic (mesonephric-like) tumors [Mirkovic et al. 2018, McCluggage 2022]. In a subset of colorectal cancers, the p.G12C-variant may be associated with a more aggressive clinical behavior [Chida et al. 2021]. In the present study, we summarized the prognostic data in correlation to different mutational sites within the KRAS-gene reported in the literature (Tab. S1, Fig. 2). Thirty cases of endometrial and only eleven cases of ovarian tumors included data on prognostic outcome in correlation to the mutational site. The lowest rates of recurrent disease and/or death of disease were reported for cases with a p.G12C-alteration and for tumors with morphologic characteristics of MLA and a KRAS-wildtype status in patients with endometrial tumors (Fig. 2). For the ovarian localization, five out of eleven patients recurred without any predilection of a KRAS-mutational site (Tab S1). There are currently not enough data to draw any valuable conclusions regarding the prognostic impact of the different mutational sites within the KRAS-gene.
In the case of Al Nabhani et al. [2022], the endometrial primary and the ocular disease depicted the same mutational event KRAS-p.G12D.
Two cases in the study of daSilva et al. [2021] showed an identical clonal KRAS-mutation in the endometrial primary and the abdominal recurrence (KRAS-p.G12V) and also within KRAS-p.G12D when a uterine MLA and its pulmonal spread were compared. An identical clonal KRAS-alteration was observed in one case of the "Leipzig cohort" in the endometrial primary and different metastatic sites (case number 5, see Table 1).
Pors et al. [2021] did not report details for mutational analyses comparing endometrial and/or ovarian primary and their recurrence or the metastatic disease.
The Kristen rat sarcoma (KRAS) gene is one of the most common mutational alterations in solid tumors [Yang et al. 2023] and occurs in approximately 22% [Forbes et al. 2011].
Oncogenic mutations of the RAS-gene are frequent in colorectal cancers, affecting about 40% of the cases, of which 85% refer to KRAS [Cherry et al. 2023], mostly at codon 12 [Li et al. 2018]. In colorectal cancer patients, a KRAS-mutation occured at p.G12V in 26.4%, followed by p.G12D (19.2%), p.G13D (16.5% [Malapelle et al. 2021]) and p.G12C in 8.5% [Schirripa et al. 2020]. In the present study, a p.G12V KRAS-alteration was most commonly seen within MLA of the female genital tract (36.5%; 44/121), followed by p.G12D (33%) and p.G12C (7.4%; for details see Fig. 2).
Recently, targeted inhibitors of KRAS have offered a breakthrough for solid tumors [Yang et al. 2023, Skoulidis et al. 2021, Fakih et al. 2022, Strickler et al. 2023, Hong et al. 2020]. Sotarasib is a specific and irreversible inhibitor of the GTPase-protein in p.G12C-KRAS-mutated cancers. Pre-treated colorectal cancer patients with progression of disease showed an objective response rate of 9.7% in a phase 2 trial [Fakih et al. 2022]. Within a phase 3 trial of 345 patients with non-small cell lung cancer (NSCLC) who progressed after previous treatment with chemotherapy and/or checkpoint inhibition, an improved progression-free survival after treatment with sotarasib was seen compared to docetaxel (5.6 versus 4.5 months, hazard ratio 0.66; p = 0.0017 [de Langen et al. 2023]). In patients with chemotherapeutically pretreated metastatic pancreatic cancer, sotarasib was associated with an objective response rate of 21% [Strickler et al. 2023]. In a basket trial of metastatic solid tumors (pre-treated with a median of three lines of chemotherapy), there was a disease control rate of 88.1% in NSCLC and of 73.8% in colorectal cancer patients [Hong et al. 2020]. Within that trial, one out of two patients with (non-mesonephric-like) endometrial carcinoma showed stable disease.
According to the present results, only a minority of MLA of the female genital tract may potentially be targetable by sotarasib because of p.G12C alteration: 7.9% (7/89) with endometrial and 6.2% (2/32) with an ovarian site (for details see Fig. 3b, c).
Until now, targeted therapy against a KRAS-mutation at p.12GC with sotarasib has only been approved for lung and colorectal cancer by the Federal Drug Adminitration (FDA) and European Medical Association (EMA) [Skoulidis et al. 2021, Fakih et al. 2022, de Langen et al. 2023]. Adagrasib is another covalent selective inhibitor targeting p.G12C [Cherry et al. 2023]. The bicyclic peptide KS-58 showed anticancer activity within the lung cancer cell line A427 and the pancreatic cancer cell line PANC-1, harboring a p.G12D-alteration [Sakamoto et al. 2020]. Furthermore, KS-58 exhibited anti-cancer activity against PANC-1 xenografts in mice in that study.
Preclinical data targeting the molecule Son of Sevenless-1 (SOS-1), catalysing the conformation of KRAS may offer a possible pan-KRAS inhibition [Kessler et al. 2021] rather than a single mutation inhibition. Different approaches for RAS-targeting are reviewed by Erlanson & Webster [2021].
Abstracting the data of MLA of the female genital tract, mesonephric-like histology per se is associated withs a poor prognostic impact. A KRAS-mutation is the most frequent molecular event and is seen in 92% of endometrial and 87% of ovarian tumors [daSilva et al. 2021]. Within the KRAS-mutated cases, a p.G12C alteration is seen in 7.9% of endometrial and 6.2% of ovarian MLA (see Fig. 3b,c and supplementary Table 1). Therefore, targeted inhibition of KRAS-G12C may offer a potential treatment approach only in the minority of MLA of the female genital tract.