Relationship between gastric xanthoma, gastric precancerous lesions, and gastric cancer: A retrospective study

To evaluate the relationship between gastric cancer and its precancerous lesions and gastric xanthoma.

0.018% to 7%. 7,8 Typical endoscopic appearances of GX include single or multiple yellow-white nodules or plaques that are round or oval in shape, with a rough surface and a clear boundary. 9 Histologically, a large number of lipid-phagocytic monocytes and macrophages transform into foam cells to form xanthoma, which are clustered in the mucosa and submucosa in the shape of nests. 10 Clinical features of GX are usually non-specific. Some patients present with dyspepsia or upper abdominal discomfort, which may be related to its concomitant diseases. 11 The diagnosis of GX relies primarily on gastroscopic findings and biopsy-proven histological evidence. In recent years, with advanced endoscopic techniques and change in environmental impact on diet and other factors, the incidence of GX tends to increase and has attracted attention. Gastric atrophy and intestinal metaplasia have been reported to be more common in patients with GX than in those without. 12,13 Moreover, GX has been found to be a risk factor for the development of gastric cancer in patients with chronic atrophic gastritis (CAG) and is expected to be a predictive indicator for gastric cancer. 14 However, comprehensive studies on the relationship between GX and gastric precancerous lesions and gastric cancer are still lacking.
Therefore, we conducted this retrospective study aiming to evaluate and further elucidate the relationship between GX and gastric precancerous lesions and gastric cancer.

| Diagnostic criteria and patient grouping
Patients were diagnosed based on gastroscopic and/or histopathological findings. All the endoscopic images were examined by two or more specialized endoscopists. Narrow-band imaging (NBI) magnifying endoscopy was performed in some of the patients for the diagnosis of intestinal metaplasia. For patients with lesions that could not be diagnosed under endoscopy, the diagnosis was further confirmed by histopathological findings. Notably, intraepithelial neoplasia and gastric cancer were confirmed by histopathological examination.
Based on their diagnosis, the patients were further divided into three groups: (a) the chronic gastritis group, including patients with CAG and chronic nonchronic atrophic gastritis (CNAG); (b) the precancerous lesion group, including patients with intestinal metaplasia and intraepithelial neoplasia; and (c) the gastric cancer group. According to the novel scoring system for gastric cancer screening, 15 patients aged ≥50 years was assigned a different score, which fully reflected the risk of gastric cancer. Therefore, we also categorized the patients into two groups according to their age: (a) <50 years; (b) ≥50 years.

| Data collection
Patients' age, sex, endoscopic and/or histopathological findings, presence and grading of bile reflux, location of GX, and the presence of single or multiple GX lesions were obtained from their medical records. The flowchart of patient enrollment in this study is shown in

| Statistical analysis
Statistical analyses were performed using SPSS software version 26.0 (IBM, Armonk, NY, USA). Continuous variables are presented as mean ± standard deviation and compared using the one-way analysis of variance (ANOVA). Categorical variables are presented as numbers and percentages and were compared between groups using the chisquare test. The multivariate logistic regression model was used for multivariate analysis which included all the variables. A two-sided P value of less than 0.05 was considered statistically significant.  Figure 2, the detection rate of GX in the gastric antrum (n = 714, 52.50%), body (n = 251, 18.46%), the entire stomach (n = 249, 18.31%), and cardia and fundus (n = 146, 10.74%) decreased gradually. It was also higher in patients aged 50 years and older compared with those younger than 50 years of age (3.95% vs 1.64%, P < 0.001) and in male patients than in female patients (3.26% vs 2.40%, P < 0.001) ( Table 1). Although the detection rate of GX showed an overall increasing trend with age (0.33% to 4.50%; Figure 3), there was no significant difference in the detection of GX among different seasons (P = 0.628) or between the presence and absence of bile reflux (P = 0.709).

| Detection of GX in different gastric mucosal lesions
As shown in Table 2, the detection rate of GX was highest in patients with precancerous lesions (n = 308, 8.39%), followed by the patients with gastric cancer (n = 71, 5.44%) and chronic gastritis (n = 981, 2.29%), respectively (P = 0.001). The chronic gastritis group was further divided into the CAG and CNAG groups for analysis. Age, the proportion of male sex, and the detection rate of GX increased with the progression of gastric mucosal lesions from CNAG (0.68%), CAG and intestinal metaplasia (4.95%) to intraepithelial neoplasia and gastric cancer (6.44%) (Table S1). In addition, we performed a further stratified analysis of the detection rate of GX in the three groups according to patient's age and sex. As shown in Table 3, the detection rate of GX in the chronic gastritis group was significantly lower than that in the precancerous lesion group and gastric cancer group in both  while among patients aged 50 years and older, the detection rate of GX in the gastric cancer group was significantly lower than that in the precancerous lesion group (P < 0.01). As shown in Table 4, in both male and female patients, the detection rate of GX in the chronic gastritis group was significantly lower than that in the precancerous lesion group and the gastric cancer group, and that in the precancerous lesion group was significantly higher than that in the gastric cancer group (all P < 0.05). Multivariate analysis was conducted to determine the clinical indicators in patients with gastric mucosal lesions at different stages. As shown in Table 5

| Clinical characteristics of patients with GX in different gastric mucosal lesion groups
We further analyzed the characteristics of GX patients with different gastric mucosal lesion, showing significant differences in patients' age and sex, bile reflux, number and location of the GX lesions among the

| DISCUSSION
Previous studies have reported that the incidence of GX is strongly associated with age, with a high incidence in those over 60 years of age, 16 and that GX is usually a solitary lesion 17 that is more commonly seen in men and in the gastric antrum. 18 In the current study, 47 736 cases were retrospectively included and GX was detected in 1360 (2.85%) patients. In addition, we found that the detection rate of GX was significantly higher in patients aged 50 years and older than in those younger than 50 years of age (3.95% vs 1.64%, P < 0.001). And it was significantly higher in male patients than in female patients (3.26% vs 2.40%, P < 0.001). In addition, GX in our patients was more likely to be a single lesion and locate in the gastric antrum, which is consistent with the previous studies.
So far, large, multicenter epidemiological surveys are lacking, most of the existing reports are single-center studies with limited samples.
The incidence of GX was reported to be 0.018% in Bulgaria in 1999. 7 In a study conducted in China in 1989, the detection rate of GX was 0.8%. 19 While in Turkey, the incidence of GX was reported to be 0.23% in 2004, which increased to 4.9% in 2016. 13,20 These data suggest that the detection rate of GX has beem increasing though there are regional differences.
The pathogenesis and risk factors of GX remain unclear. Some studies suggest that abnormal lipid metabolism can lead to the transfer of lipoprotein from gastric mucosal capillaries to mucosal or submucosal intercellular space, and to form oxidized lipoprotein under the action of oxygen free radicals. Chemotaxis of monocytes or macrophages is observed, which engulfs oxidized lipoproteins and other lipids, and the cells turn into foam cells; xanthoma is formed when foam cells gather in large numbers. 10,21 Some researchers also believe that local tissue damage of gastric mucosa leads to the accumulation of lipids in mucosal epithelial cells, which is involved in the formation of GX. In 1996, H. pylori was first detected in GX tissue samples, 22 and some studies found that the infection rate of H. pylori in patients with GX was significantly higher than that in those without. 23 H. pylori infection is the main pathogenic factor of many gastric diseases including chronic gastritis, peptic ulcer, and gastric cancer. GX has been suggested to be related to chronic inflammation of gastric mucosa, 20 hyperplastic polyps, 24 cholestasis, 25 and immunosuppression. 26 It is concluded that GX may be related to gastric cancer and precancerous lesions. Yamashita et al revealed that the detection rate of GX was significantly higher in patients with gastric cancer than in those without, and further multifactorial analysis showed that GX was an independent risk factor for gastric cancer. 27 Shibukawa et al and Yan et al both reported that GX was a risk factor for early gastric cancer after H. pylori eradication. 28,29 And Kitamura et al found that GX was strongly correlated with the differentiation of gastric cancer. 30 Xiao et al found that the incidence of gastric atrophy and intestinal metaplasia was significantly higher in patients with GX than in those without GX and that age and intestinal malformation were independently associated with GX. 12 These findings have all shed some light on the close association between GX and gastric mucosal injury, gastric precancerours lesions, and gastric cancer, but many studies included a small sample size and comprehensive studies of the relationship between GX and chronic gastritis, precancerous lesions, and gastric cancer are lacking.
In our study, the patients were divided into the chronic gastritis, precancerous lesion, and gastric cancer groups, and the detection rate of GX was analyzed. We found that the detection rate of GX in the chronic gastritis group was significantly lower than that in the precan- Some studies suggest that GX does not need treatment, 31 but requires endoscopic monitoring during the follow-up period. 32 We further analyzed the proportion of single GX lesion in each group, and found that in the chronic gastritis group the proportion of single GX lesion was significantly higher than that of multiple GX lesions. The proportion of multiple precancerous lesions was significantly higher than that of single lesion. There was no significant difference in the proportions of single and multiple GX in the gastric cancer group. It can be concluded that in chronic gastritis, GX tends to be solitary, and when gastric disease progresses GX tends to be multiple lesions. With the progress of gastric disease, GX becomes more complex.
There were some limitations to our study. First, this was a retrospective study; therefore, clinical data of the included patients were not comprehensive. For example, data of H. pylori infection, lifestyle, and genetic background was missing. And due to the lack of endoscopic images of the whole stomach, the presence of lesions including GX might have been missed, leading to a low detection rate. Second, this was a single-center study including patients with limited geographic origin and dietary structure. Third, some patients with intestinal metaplasia were diagnosed by NBI magnifying endoscopy, which might have led to biases because a small proportion of patients might have been classified as having chronic gastritis. Therefore, multicenter prospective studies with large sample sizes are needed to elucidate the causal relationship between GX and gastric cancer and its precancerous lesions in order to provide more targets and indicators for the prevention and early diagnosis of gastric cancer.
In conclusion, our study confirmed a high detection rate of GX in patients with gastric cancer and precancerous lesions and their relationship. Currently, a large-sample, multicenter, prospective study is being conducted by our group to further confirm whether GX is an independent risk factor for gastric cancer and precancerous lesions.
To prevent the development of gastric cancer and precancerous lesions, physicians should pay more attention to patients with GX. If GX is identified by endoscopy, mucosal changes, in particular precancerous lesions, needs to be carefully observed, and close follow-up is required even if no gastric cancer is found.