We encountered a case of a late recurrence of gastric cancer at POY 12. The histopathological findings were useful in confirming the diagnosis of postoperative recurrence of gastric cancer. The histology of the recurrent tumor indicated the proliferation of adenocarcinoma cells, similar to the histology of the previous surgical specimen. Gastric adenocarcinomas demonstrate either a CK7 or CK20 expression pattern; the recurrent lesion was CK7 positive/CK20 negative, consistent with the expression pattern of the primary lesion. No epithelial tumor was detected in esophagogastroduodenoscopy and colonoscopy, and other cancers were not suspected based on contrast-enhanced CT and PET results. These findings further support the diagnosis of recurrence of gastric cancer.

To consider the reasons for late recurrence, we investigated the expression of CD44v9 with immunohistochemical staining. CD44 is a marker for cancer stem cells (CSCs) found in various solid tumors, such as breast, colon, and gastric cancers. It is a major adhesion molecule in the extracellular matrix. [7] CSCs, a unique subpopulation in tumors, can initiate tumor growth and sustain tumor self-renewal [8]. CD44 gene transcripts undergo complex alternative splicing, among which CD44v9 expression is significantly associated with poor prognosis concerning overall survival and recurrence-free survival [9,10,11,12,13]. CD44v9 stabilizes the glutamate–cystine transporter and promotes the uptake of cystine, which is required for glutathione (GSH) synthesis in the cell [14]. GSH is cells’ most abundant non-enzymatic antioxidant molecule and directly removes intracellular reactive oxygen species (ROS) [14]. GSH peroxidase 2, the gastrointestinal form of GSH peroxidase, is an antioxidant enzyme that catalyzes the reduction of intracellular ROS using GSH as a reductant. This defense mechanism against ROS prevents apoptosis of CSCs, making them resistant to anticancer drug therapy and radiotherapy [15]. Conversely, silencing CD44v9 with siRNA has been reported to promote apoptosis and cell cycle arrest and inhibit cell proliferation [16]. In our case, the expression of CD44v9 was attenuated in the recurrent lesion compared to the primary lesion (Fig. 5a, b). The defense mechanism against ROS in the recurrent lesion may be weakened by the decreased expression of CD44v9. When the number of CD44v9-positive cells is low, tumor formation and cell proliferation may also be suppressed. The competing activities of cell proliferation and apoptosis may contribute to tumor growth inhibition. Meanwhile, the Ki-67 labeling index of < 3% (Figs. 1e, 4d) and the weak accumulation of fluorodeoxyglucose by PET suggest that the tumor cells are in a state of cell cycle arrest or dormancy [17]. Therefore, the expression of CD44v9 in recurrent lesions may cause late recurrence.

We hypothesized that the degree of CD44v9 expression in recurrent lesions correlates with the recurrence time. Existing reports suggested that CD44v9-positive tumors have a higher recurrence rate and poor survival than CD44v9-negative tumors [9,10,11,12,13]. On the other hand, no papers have examined the expression of CD44v9 in recurrent tumors. We investigated the expression of CD44v9 in the primary and recurrent lesions in seven patients who underwent radical gastric cancer surgery and resection of recurrent lesions in our hospital. Among them, CD44v9 expression was positive in three cases. Case 1 was the present case of gastric cancer with para-aortic lymph node recurrence 12 years postoperatively, Case 2 was gastric carcinoma with para-aortic lymph node recurrence 5 years postoperatively, and Case 3 was esophagogastric junction carcinoma with skin metastasis 8 months postoperatively. The expression of CD44v9 was classified as primary/recurrence: positive/slightly positive in Case 1, positive/negative or unclear in Case 2, and positive/positive in Case 3. In Case 3 (early recurrence), high expression of CD44v9 in the recurrent lesion was observed. We suspect that in gastric cancers expressing CD44v9, cases with a lower percentage of CD44v9 expression in the recurrent lesion have a longer time to recurrence than those with a higher percentage of CD44v9 expression.

A limitation of our study is that this conclusion is drawn from a small number of cases. There are only a few cases in which surgery for gastric cancer recurrence is indicated, and consequently, pathological findings in recurrent lesions are rarely reviewed.

A defense mechanism against ROS has been reported as a mechanism that causes the recurrence of CD44v9-positive gastric cancer. Consequently, CD44v9-positive gastric cancer grows in metastatic organs, repeatedly self-renews, and proliferates to form recurrent lesions. In the present case, the degree of CD44v9 staining in recurrent lesions was suggested to be related to the recurrence time.