Ovarian Combined Serous Borderline Tumor/Low-grade Serous Carcinoma and Mesonephric-like Lesion: Report of 2 Cases With New Observations.

The second case was a 58-year-old woman who presented with bilateral ovarian masses. She underwent TH-BSO, omentectomy and staging biopsies. Gross examination revealed a 30 × 23 × 14 cm unilocular cystic lesion in the left ovary and a 17 × 15 × 11 cm multilocular cystic mass in the right ovary. Similar to case 1, both ovaries exhibited papillary excrescences on their external surface. Histological examination revealed conventional type SBT/APST in both ovaries. In addition, the left ovarian tumor demonstrated a few areas of glandular proliferations composed of small tubules lined by cuboidal cells with a moderate degree of cytologic atypia and eosinophilic colloid-like secretions in the lumen ( Figures 4A and 4B ). Each area measured less than 5 mm in greatest dimension. This component, classified as mesonephric-like tubules, displayed some hyperplastic features that were considered insufficient to establish a diagnosis of MLA. In particular, neither glandular confluence nor desmoplastic reaction were present. In some areas, individual mesonephric-like glands and a few glandular clusters were seen in close proximity to the serous-type epithelium ( Figure 4G ). The epithelial cells in the mesonephric-like glands displayed mild to moderate atypia and lacked cilia. Non-invasive implants were found in the right pelvic sidewall, but evidence of invasive carcinoma or extra-ovarian benign/hyperplastic mesonephric-like proliferation was not identified. Similar to the first case, both the SBT/APST and the mesonephric-like tubules displayed focal/patchy positivity of p16 and a wild-type pattern of p53 staining. The former component was negative for Gata3 ( Figure 4C ), diffusely positive for WT-1 ( Figure 4D ) and ER ( Figure 4E ), and focally positive for PR, and while the mesonephric-like component was focally positive for Gata3 ( Figure 4C ) and negative for WT-1 ( Figure 4D ), ER ( Figure 4E ), and PR. A luminal pattern of CD10 staining was present in the mesonephric-like tubules but not in the SBT/APST ( Figure 4F ). Individual mesonephric-type glands and adjacent serous-type epithelium ( Figure 4G ) displayed a distinct immunoprofile for each component ( Figures 4H : Gata3; 4I : WT-1).

Six cases of ovarian combined SBT/LGSC with MLA and 1 case of SBT/LGSC with co-existent mesonephric-like carcinosarcoma have been reported in the literature [ 7 , 15 – 18 ]. Clinicopathologic features are summarized in Table 1 and immunohistochemical and molecular findings are summarized in Table 2 . The patients with mesonephric-like tumor, including 7 published cases and 1 case from this report, ranged in age from 61 to 80 years (mean, 67; median, 64). All 8 cases with a malignant mesonephric-like component, with or without co-existing LGSC, developed metastatic disease involving the abdomen/pelvis. Other sites involved by metastatic tumor included liver, lung, and lymph nodes (pelvic and intrathoracic). Based on the morphologic findings and molecular changes, two models were proposed to explain the histopathogenesis of the mixed tumors in our report as well as in other studies ( Figures 5A and 5B ).

This report documents two cases of mixed serous tumor and mesonephric-like lesions in the ovary. To the best of our knowledge, our case 1 represents the eighth reported case of ovarian malignant mesonephric-like tumor coexisting with SBT/LGSC. Strikingly, all these tumors display similar histomorphology, immunophenotype, molecular alterations, and clinical behavior. The serous component of case 1 was composed of both cribriform SBT/niLGSC and conventional SBT/APST, rather than a pure conventional borderline tumor. For micropapillary/cribriform type SBT, we favor designation as niLGSC since women with niLGSC are more likely to develop invasive serous carcinoma than women with conventional SBT/APST [ 19 , 20 ]. Consistently, a component of LGSC, either non-invasive or invasive, was present in 4 of 7 reported cases ( Table 1 ). Immunohistochemically, this component was characterized by WT-1 and ER/PR expression and lack of expression of mesonephric markers Gata3, TTF-1 and CD10 ( Table 2 ). Interestingly, the median age of the patents with these mixed tumors was 64 years, which is similar to those reported in MCs and pure MLAs, but decade older than those of pure SBT/APST or LGSC [ 20 ]. The other component in the biphasic lesion of case 1 exhibited typical features of MLA, illustrated by a mixed growth pattern, intraluminal eosinophilic secretions and a Gata3-positive, WT-1/ER/PR-negative immunophenotype. Similar to other reported cases, both components in our case showed patchy p16 expression and wild-type p53. Although only 5% of tumor was MLA in the ovary, this component accounted for more than 95% of the metastatic tumor in the omentum. A similar observation has been reported in other cases [ 16 , 18 ]. Our case—as well as reported cases with a malignant mesonephric-like component, with or without co-existing LGSC—all developed metastatic disease involving the abdomen/pelvis and some cases had spread outside of the abdominal cavity, indicating their aggressive behavior. At a molecular level, NGS and Sanger sequencing analysis revealed a common KRAS G12V driver mutation in both the SBT/niLGSC and MLA components in case 1. In fact, since its first description [ 16 ], there are several published reports with molecular analysis confirming a clonal origin for these mixed tumors ( Table 2 ). The identical activating KRAS mutations, including G12D [ 15 ], G12C [ 17 ], and G12V (case 1), and NRAS Q61R mutation [ 7 , 16 ] have been reported in both SBT/LGSC and MLA components. Some reported cases had mixed histology but it was impossible to dissect the distinct components entirely due to the intimate association [ 7 ]. These cases, in which both SBT/LGSC and MLA components were dissected and sequenced together, demonstrated a KRAS G12V mutation in one case and a NRAS Q61R in the other. In addition to common KRAS and NRAS mutations, other shared genetic alterations have been reported in both components, including a PIK3CA p.E545K mutation, gains in chromosome 1q and 18p and losses in chromosomes 1p, 4, 18q, and 22. It is conceivable that, despite having a common clonal origin with shared genetic alterations, there must be other distinct genetic changes to explain their distinct phenotypes in a mixed tumor. Consistent with this assumption, additional private aberrations were detected either only in SBT/LGSC ( KDM5A and STAG2 ) or only in MLAs ( BCOR , AMER1, MYCN, HIST1H3I , gains of chromosomes 6p and 17) [ 7 , 16 , 17 ]. The literature indicates that KRAS somatic mutations are the most common genetic alteration of both ovarian (up to 87%) and endometrial (up to 92%) MLAs [ 7 ]. These mutations were also detected in our case 1 and 3 of 6 reported cases which had molecular analysis. Ras/Raf/MEK/MAPK pathway alterations are postulated to be drivers of both MCs and MLAs [ 7 ]. Consistently, it has been demonstrated that MCs or MLAs that lack KRAS mutations may harbor hotspot mutations in other RAS/RAF family genes such as NRAS or BRAF . However, compared with KRAS mutation, NRAS or BRAF mutations in these tumors are very rare – only one pure cervical MC with NRAS mutation [ 9 ] and one pure endometrial MLA with BRAF mutation [ 7 ] have been reported. Several studies demonstrated that NRAS is a critical oncogenic driver in the progression of SBT/APSTs to LGSCs. In one study, 5 of 58 (9%) invasive tumors with adjacent SBT harbored activating NRAS mutations [ 21 ]. Another study demonstrated that NRAS mutations were detected in 26.3% of LGSCs, but none were detected in the SBT/APST cohort [ 22 ]. Similarly, our previous study showed that NRAS Q61R mutations were detected in 2 of 56 (3.6%) invasive LGSCs but not in any of the SBT/APSTs or niLGSCs [ 23 ]. Interestingly, while NRAS p.Q61R driver mutations are not common in either LGSCs or MC/MLAs, 3 of 7 mixed MLAs and SBT/LGSCs with molecular testing had this mutation. It is conceivable that acquisition of an NRAS mutation—similar to, or perhaps more effectively than, a KRAS mutation—defines the invasive nature of SBT/APST, causing its progression into a malignant tumor with distinct serous and mesonephric lineages [ 16 ]. In case 2, we report a conventional type SBT/APST with mesonephric-like differentiation/hyperplasia, a finding which has not been previously described. The predominant component was SBT/APST in which the tumor cells were positive for WT-1, ER, and PR, but negative for Gata3, TTF-1, and CD10. In this 30 cm tumor, few microscopic foci of small tubular clusters, each measuring less than 5 mm, also were present. These small tubules displayed typical mesonephric morphology and immunophenotype with positive Gata3 expression and a luminal pattern of CD10 staining. The degree of glandular crowding was consistent with hyperplasia but considered insufficient to establish a diagnosis of MLA. Although a similar phenomenon has not been reported in any serous type lesions, a recent case series described a case of ovarian endometrioid borderline tumor where the glands focally exhibited “mesonephric-like” differentiation [ 13 ]. It has been well accepted that at least some MLAs, if not all, arise from a Müllerian origin. In a tumor with mixed serous and mesonephric morphology, the presence of SBT/APST with an associated biphasic invasive carcinoma suggests a Müllerian origin for the entire malignant process. In theory, a Müllerian-type progenitor cell can acquire KRAS or NRAS mutations and develop into an SBT/APST that continues to give rise to both invasive LGSC and MLA ( Figure 5A ). Alternatively, it is also plausible that the tumor with mixed histology may originate from a pluripotent stem cell in the embryonic ridge which acquires KRAS or NRAS mutations and then differentiates in parallel into both Müllerian and mesonephric lineage ( Figure 5B ) [ 16 ]. The presence of a non-malignant mesonephric-like component in case 2 indicates the latter possibility does theoretically exist. It has been demonstrated that, in serous type lesions, endosalpingiosis is frequently associated with SBT/APST and even harbors the same KRAS/BRAF mutation as the ovarian tumor despite its benign morphology [ 24 ]. Likewise, as the precursor lesion, the individual mesonephric-like glands in case 2 may represent an analogous phenomenon to endosalpingiosis and may already contain KRAS or NRAS mutation, although this is purely speculative. In summary, we report two cases of ovarian combined SBT/LGSC and MLA or mesonephric-like differentiation/hyperplasia. In concordance with previously reported cases, case 1 contained identical KRAS mutations in both tumor components. These findings provide further evidence to demonstrate the clonal relationship between these morphologically and immunophenotypically distinct components. It also supports the theory that, unlike cervical MCs originating from mesonephric remnants, MLAs are derived from Müllerian-type lesions which differentiates into the mesonephric lineage. The presence of a hyperplastic mesonephric-like lesion/differentiation in case 2 indicates that a precursor lesion in the same lineage with the potential to develop into MLA exists in the ovary.

Mesonephric-like adenocarcinomas (MLAs), commonly occurring in the uterine corpus and ovary, are rare malignant neoplasms displaying mesonephric differentiation [ 1 ]. As newly described entities, MLAs have been added to the recent 2020 World Health Organization (WHO) Classification of Female Genital Tumors [ 2 ]. Despite the absence of obvious mesonephric/Wolffian-type precursor lesions, MLAs exhibit similar histological and immunohistochemical features as well as molecular alterations to those of mesonephric carcinomas (MCs) of the uterine cervix which is thought to arise from mesonephric remnants [ 3 – 5 ]. Both MLAs and MCs feature a variety of morphologies including tubular, glandular/pseudoendometrioid, ductal, retiform, papillary or solid patterns. Small glands and tubules with eosinophilic intraluminal secretions are frequently seen. These tumors are characterized by nuclear expression of Pax8, Gata3 and TTF-1, luminal staining of CD10, lack of ER/PR expression, and a wild-type p53 staining pattern [ 6 ]. At a molecular level, KRAS somatic mutations has been reported in a high proportion of both MLAs and MCs, with up to 89% in the former and 100% in the latter [ 7 – 11 ]. Although the cell of origin of primary MLAs is unknown, the literature indicates that at least some of these tumors are derived from transdifferentiation of Müllerian-type lesions into those with Wolffian/mesonephric lineage. Indeed, some of these tumors are associated with Müllerian-type lesions in the ovary including endometriosis [ 12 – 14 ], serous cystadenoma [ 15 ], endometrioid borderline tumor [ 13 ], and serous borderline tumor/low-grade serous carcinoma (SBT/LGSC) [ 7 , 15 – 17 ]. Thus far, 7 cases of ovarian combined SBT/LGSC and MLA [ 7 , 15 – 17 ] or mesonephric-like carcinosarcoma [ 18 ] have been reported and the presence of identical KRAS , NRAS , or PIK3CA mutations in both components provides convincing evidence that these distinct components are clonally related. On the other hand, the presence of private mutations in each component indicates lineage-specific differentiation with distinct morphology and immunophenotype. Here, we report two cases of ovarian combined serous tumor and mesonephric-like lesion: one is a mixed SBT/LGSC and MLA, and the other is an SBT with focal mesonephric-like differentiation/hyperplasia. The observations in the latter case have not been reported previously.