This case highlights two important clinical issues. First, gastric PBL can present as multiple polypoid lesions in HIV- and EBV-negative patients. Second, gastric PBL may coexist with gastric adenocarcinoma. To the best of our knowledge, this is the first case presenting endoscopic images of multiple polypoid lesions of gastric PBL and reporting the coexistence of PBL with gastric adenocarcinoma.

First, gastric PBL can exhibit multiple polypoid lesions in HIV- and EBV-negative patients. Although PBL of the stomach was first reported in 1998 [9], only a few reports have been published. To our knowledge, 11 cases of gastric PBL were reported from 1998–2023 [9,10,11,12,13,14,15,16,17,18,19]. The clinical and imaging characteristics of the published cases are summarized in Table 1. Seven of the 11 reports described endoscopic findings, and three of these included endoscopic images. Four of the six cases showed ulcerative lesions, while threes exhibited polypoid masses or irregularly raised lesions; however, no endoscopic images were available. This is the first report of endoscopic imaging of multiple polypoid lesions in gastric PBL.

The endoscopic findings of gastric lymphoma vary. Primary gastric lymphoma is mostly a B-cell non-Hodgkin’s lymphoma, with mucosa-associated lymphoid tissue (MALT) lymphoma and DLBCL [20]. The endoscopic appearance in MALT lymphoma tends to appear as a superficial type, whereas gastric DLBCL is a mass/polypoid or ulcerative type, and these lesions are often multiple [21, 22]. The endoscopic findings of gastric PBL may be similar to those of DLBCL, as previous reports have shown ulceration and raised/polypoid lesions. Notably, gastric PBL appears to exhibit a predilection for the upper to middle part of the stomach, with 8 out of 12 reported cases, including the present case (three cases did not report the location), showcasing lesions in the fundus and body of the stomach (Table 1). In addition to these findings, further cases are needed to clarify the endoscopic features of gastric PBL.

HIV and EBV have been implicated in the development of PBL [2, 3]. However, they are not necessarily involved in PBL of the gastrointestinal tract. In a study of 590 PBL cases, 43 out of 83 PBL cases (52%) occurring in the gastrointestinal tract were HIV-positive [6]. In a study of PBL in the intestinal tract, 6 of 17 cases (27%) were HIV-positive, and 9 of 16 cases (56%) were EBV-positive [23]. In a case series of gastrointestinal PBL in HIV-negative cases, 19 of 31 (61%) cases were EBV-positive [7]. Regarding gastric PBL, 3 of 11 (27%) cases were HIV-positive and four of nine (44%; two cases did not report EBV status) were EBV-positive (Table 1). Both HIV and EBV tested negative in this case, and their involvement in the development of gastric PBL may not have been significant. This suggests the need to include PBL in the differential diagnosis of gastrointestinal tract lymphomas, even in patients without HIV or EBV infection.

PBL is sometimes difficult to diagnose pathologically due to its rarity, especially when it occurs in the extraoral region of HIV- or EBV-negative patients. PBLs mainly have cytomorphological features, such as large plasma cells or large immunoblasts expressing plasma cell markers and lacking B-cell markers [6, 24]. Differential diagnoses of PBL include plasmablastic myeloma, DLBCL (not otherwise specified), and undifferentiated carcinoma. In this case, plasmablastic myeloma was excluded because of the absence of features favoring myeloma, such as monoclonal paraproteinemia and hypercalcemia. In addition, the differentiation between DLBCL and undifferentiated carcinoma from PBL typically relies on the expression of pan-B-cell markers and cytokeratin, respectively. This patient had an unusual phenotype with partial expression of B-cell marker (CD20) and cytokeratins (AE1/AE3 and CAM5.2), which led to difficulties in the initial diagnosis [25]. Furthermore, the combination of CD20 and PAX5 expressions is useful in classifying DLBCL and PBL, with negative or weakly positive CD20 and PAX5 being characteristic phenotypes of PBL [26]. In this case, subsequent analysis revealed negative PAX5 expression in the tumor cells (Supplementary Fig. 1), combined with weak CD20 expression, indicating the diagnosis of PBL.

The second clinical suggestion is that gastric PBL can coexist with gastric adenocarcinoma. Gastric lymphoma and gastric cancer are rarely diagnosed simultaneously. A previous Japanese study reported that 4 of 121 (3.3%) primary gastric lymphomas resected surgically coexisted with gastric adenocarcinoma [27]. In a multicenter study in China, 5% (24/474) of patients with gastric lymphoma were diagnosed with gastric cancer, and only one of these patients was synchronously diagnosed with lymphoma and gastric cancer [28]. Most of the lymphomas in these studies were DLBCL or MALT lymphomas. Given the rarity of gastric PBL, reports on its complication with gastric cancer are absent. Differentiating gastric PBL from other lymphomas and diagnosing synchronous gastric cancer based on endoscopic appearance can be difficult. In this case, the second biopsy revealed gastric PBL, and the coexistence of gastric cancer was confirmed. Repeated pathological examinations contribute to the accurate diagnosis of rare lymphomas and the detection of synchronous gastric cancer.

The primary etiology of gastric adenocarcinoma and MALT lymphoma is Helicobacter pylori (H. pylori) infection [29, 30]. Moreover, H. pylori has also been implicated in some gastric DLBCL [31]. Precancerous conditions resulting from chronic gastritis caused by H. pylori contribute to the synchronous or metachronous occurrence of gastric cancer and lymphoma [28]. However, the role of H. pylori infection in gastric PBL development remains unclear. Previous studies have identified HIV [2], EBV [6], organ transplantation [5], chronic inflammatory disease [32], and immunosenescence in the elderly [5, 33] as factors associated with PBL. In our case, age was the sole relevant factor, with no history of immunodeficiency or immunosuppressive treatment. In this case, the patient had positive serological test results for H. pylori. As with other gastric lymphomas, the association between gastric PBL and H. pylori should be investigated in future studies. Furthermore, EBER-ISH was negative in PBL cells but positive in gastric adenocarcinoma in this case (Supplementary Fig. 2). EBV-associated gastric cancer is characterized by undifferentiated adenocarcinoma with lymphocyte infiltration in the stroma, and a relatively low frequency of lymph-node metastasis has been reported [34, 35]. The present case did not have the characteristic histology of EBV-associated gastric cancer, and the clinical significance of EBER positivity in intramucosal differentiated adenocarcinoma is unknown; however, the etiology of PBL and gastric adenocarcinoma in this case was suggested to be different.

The treatment for gastric PBL has not yet been established. In previous reports, a relatively high proportion of patients were treated with systemic chemotherapy, such as CHOP, because of the advanced stage of the disease; however, the prognosis was poor. Surgical resection was performed in two patients, both with postoperative chemotherapy (ProMACE-cytaBOM and CHOP), and one patient survived 19 months after diagnosis (Table 1). In this case, the gastric PBL was confined to the stomach, and the gastric cancer was an intramucosal adenocarcinoma without external metastasis; therefore, surgical gastrectomy seemed appropriate. Chemotherapy regimens in the postoperative period and relapse should be discussed in future.

In conclusion, our case revealed some endoscopic features of gastric PBL and suggested the rare possibility that gastric PBL may coexist with adenocarcinoma. Repeated pathologic examinations contribute to the accurate diagnosis of rare lymphomas and the detection of synchronous gastric cancer.