Plexiform Fibromyxoma with MALAT1–GLI1 Fusion with Limited Myxoid Stroma, Aberrant KIT Expression, and Diffuse D2-40 Expression: A Case Report

1. Introduction

Plexiform fibromyxoma (PFM) is a rare benign gastric mesenchymal neoplasm first described by Takahashi et al. in 2007 as “plexiform angiomyxoid myofibroblastic tumor” [1]. The term “plexiform fibromyoma” was proposed by Miettinen et al. in 2009 [2] and then PFM was officially listed as a distinct type of gastric mesenchymal neoplasm by the World Health Organization (WHO) in 2010 [3]. In the English literature, more than 130 cases have been reported [4]. PFM typically presents a multinodular plexiform growth of bland spindle cells embedded in myxoid or fibromyxoid stroma with increased vascularity [3]. Molecular studies have identified fusions of recurrent metastasis associated lung adenocarcinoma transcript 1 (MALAT1):: glioma-associated oncogene homologue 1 (GLI1) that drive GLI1 overexpression in a subset of PFM [5,6]. Here, we present the case of a cellular form of PFM with limited myxoid matrix, showing diffuse D2-40 expression, resulting in a diagnostic challenge.

2. Case Presentation

A 59-year-old woman was referred to our hospital for evaluation of a gastric submucosal tumor at the antrum that was detected at a medical checkup, and which has been under surveillance for several years. She has no remarkable past medical history or family history. Serial endoscopy revealed an enlargement of the submucosal tumor from 10 mm to 15 mm over the preceding 12 months. The laboratory data were unremarkable. Contrast-enhanced computed tomography demonstrated a well-circumscribed exophytic mass along the greater curvature and no evidence of metastatic lesion was noted. Microscopic examination of an endoscopic biopsy specimen suggested a benign or low-grade spindle cell neoplasm, i.e., gastrointestinal stromal tumor (GIST) or myopericytomatous tumor, but failed to establish a definitive diagnosis. Local resection of the stomach was performed using laparoscopic and endoscopic cooperative surgery. Postoperative course was uneventful and the patient was discharged on postoperative day 7. No evidence of recurrent tumor was identified at the time of writing (8 months after surgery).

2.1 Pathological Findings

Gross examination revealed a lobulated, solid submucosal tumor, measuring 15 × 11 × 10 mm, presenting as a protruding lesion (Figure 1). The cut surface was homogeneous without hemorrhage or necrosis.

Histologically, the tumor had a plexiform appearance and tongue-like extension into the muscularis propria, forming a lobulated mass, which was composed of bland-looking spindle cells without significant nuclear atypia, arranged in fascicular or streaming patterns with only limited fibromyxoid or myxoid stroma (Figure 2). The mitotic count was 11 per 50 high-power fields (HPF). No necrosis or lymphovascular invasion was identified, and all surgical margins were negative.

Immunohistochemical analysis indicated the tumor cells were focally positive for KIT (Figure 3a), but negative for DOG1, making a diagnosis of GIST unlikely. Succinate dehydrogenase (SDH)-deficient GIST was also excluded because SDHB expression was retained. Other differential diagnoses, including schwannoma, leiomyoma, glomus tumor, gastroblastoma, inflammatory myofibroblastic tumor (IMT), inflammatory fibroid polyp, and perivascular epithelioid cell tumor (PEComa), were also excluded based on a combination of morphology and results of extensive immunohistochemical workup (Table 1). Of note, the tumor cells stained positive for smooth muscle actin (SMA), CD10, and D2-40 (Figure 3b-e). However, cytokeratin (AE1/AE3), CD34, desmin, h-caldesmon, synaptophysin, S100 protein, HMB45, Melan A, and anaplastic lymphoma kinase (ALK) were negative. Diffuse D2-40 expression made mesothelioma and follicular dendritic cell sarcoma diagnostic concerns, but mesothelial markers (calretinin and WT-1), and follicular dendritic cell markers (CD21, CD23, and ICOS) were negative. The Ki-67 labeling index was 5.2% (91/1736) at the hot spot (Figure 3f).

Finally, because the diagnosis of PMF remained a serious diagnostic consideration, the analysis of a next-generation sequencing panel (Archer FusionPlex Sarcoma Panel) supported by the consultation system of the National Cancer Center (Tokyo, Japan) was requested. This revealed a MALAT1::GLI1 fusion, indicating a diagnosis of PFM. Although the mitotic activity appeared higher compared with that in a previous series of PFM cases, our case was considered not to fit the morphological characteristics of a malignant epithelioid tumor with GLI1 rearrangement [4,7,8].

3. Discussion

PFM is characterized by a distinctive plexiform growth pattern composed of bland spindle to stellate cells embedded within a myxoid or fibromyxoid stroma, often accompanied by vascularity [3]. This morphological appearance frequently overlaps with other gastrointestinal stromal tumors, such as myxoid GIST, nerve sheath tumor, smooth muscle tumor, glomus tumor, gastroblastoma, IMT, and inflammatory fibroid polyp [2,4,5]. PFM presents a significant diagnostic challenge because of its rarity and a variety of mimics, particularly when using small biopsy samples. In the current case, cellular morphology with only limited myxoid stroma and focal KIT expression were also a source of diagnostic confusion.

The molecular features of PFM have been investigated [4,5]. PFM lacks KIT, platelet-derived growth factor-A, and SDH gene mutations that are definitional alterations of GISTs. In 2016, Spans et al. [6] reported that a minor subset of PFM (3/16 cases, 18%) harbored MALAT1::GLI1 gene fusions. Banerjee et al. [10] reported PTCH1 inactivation via gene or chromosomal deletion in 2 of 8 cases of PFM (5 cases used for NGS analysis).

It should be noted that MALAT1::GLI1 fusion has been demonstrated in other tumors, including gastroblastoma and malignant epithelioid tumor with GLI1 rearrangement, which is a recently described epithelioid neoplasm of the soft tissue, typically with S100 protein expression and the variable expression of cytokeratin [7]. In the gastrointestinal tract, only one case involving the jejunum has been reported [8]. Although the mitotic activity and Ki-67 labeling index in our case were relatively higher (11/50 HPF and 5.2%, respectively) compared with those in a previous series of PFM cases (median: 1/50 HPF and less than 2%, respectively) [2,11], the immunophenotype, i.e., negative staining for S100 and cytokeratin, and absence of epithelioid morphology, distinct nuclear atypia, necrosis, and lymphovascular space invasion justified the diagnosis of PFM. Although the follow-up period was limited, there was no evidence of recurrent disease 3 months after resection.

A diagnosis of PFM depends on morphology and immunophenotyping to exclude other mesenchymal tumors, especially GIST. To date, no specific IHC marker for the diagnosis of PFM has been established. The utility of GLI1 IHC has been questioned [5], because MALAT1::GLI1 fusion has been identified in a minor subset of PFM. Hu et al. [11] investigated the immunohistochemical profile of 10 cases of PFM and demonstrated that all cases were positive for vimentin and SMA, and some cases were positive for CD10 (5/10), desmin (5/10), h-caldesmon (6/10), and progesterone receptor (6/10), whereas CD34, S100 protein, estrogen receptor, anaplastic lymphoma kinase (ALK), KIT, and DOG1 were all negative.

The current case highlights the diffuse D2-40 expression in PFM. D2-40 is a monoclonal antibody directed against podoplanin, a transmembrane glycoprotein involved in lymphangiogenesis, cell migration, and epithelial–mesenchymal transition [12]. Our literature search revealed only three single case reports of PFM referring to D2-40 expression. Of these three cases, one showed diffuse expression [13], whereas the remaining two cases were negative [14,15], although, unfortunately, these three case reports did not investigate MALAT1::GLI1 fusion. D2-40 IHC is a commonly used diagnostic tool among pathologists to identify lymphatic vessels, and tumors that could be differential diagnosis of PFM were usually negative for this marker (Table 1); therefore, it might be useful for the diagnosis of PFM. Future studies should examine the utility of D2-40 involving a series of PFM cases with confirmed MALAT1::GLI1 fusion.

4. Conclusions

PFM can be a diagnostic challenge because of the limited or absence of a myxoid background and focal non-specific KIT expression. Diffuse D2-40 expression might be a surrogate for MALAT1::GLI1 fusion when establishing a diagnosis of PFM in an appropriate clinicopathologic context. In this regard, the spectrum of morphology, as represented by the current case, is an essential component for avoiding diagnostic misinterpretations. Further investigations are expected to reveal the diagnostic utility and biological significance of D2-40 expression in PFM.