Circulating methylated THBS1 DNAs as a novel marker for predicting peritoneal dissemination in gastric cancer

Abstract Objectives Thrombospondin 1 (THBS1) is known to play a key role in tumor metastasis, and aberrant DNA methylation is one of the mechanisms regulating THBS1. The present study investigated whether methylated THBS1 in circulating cell‐free DNA from preoperative peritoneal lavage fluid (PPLF) and peripheral blood could be used as a potential biomarker for predicting peritoneal dissemination in gastric cancer (GC) patients. Methods The status of THBS1 methylation was detected by quantitative methylation‐specific PCR (MSP) in tumor tissues, paired PPLF, and serum from 92 GC patients. The correlation between methylated THBS1 levels and peritoneal dissemination of GC was studied, and its diagnostic value for predicting peritoneal dissemination was clarified by the receiver operating characteristic (ROC) curve. Results Aberrant THBS1 methylation in tumor tissues was significantly higher than that in paracancerous normal tissues (p < 0.0001). No THBS1 methylation was found in 40 healthy controls, and partial methylation was detected in 3 of 48 patients with chronic non‐atrophic gastritis. The frequency of THBS1 methylation in pairing PPLF and serum from 92 GC patients was 52.2% (48/92) and 58.7% (54/92), respectively. The results of methylated THBS1 in pairing PPLF and serum were similar to those of tumor tissues. Aberrant THBS1 methylation in tumor tissues and pairing PPLF or serum was closely related to peritoneal dissemination, tumor progression, and poor prognosis (all p < 0.0001). Conclusion Circulating methylated THBS1 DNAs in PPLF/serum may predict peritoneal dissemination, a potential poor prognostic factor for GC patients.


| INTRODUC TI ON
Gastric cancer (GC) is one of the most common gastrointestinal malignancies, severely threatening human health. According to the latest global cancer burden date of the World Health Organization's International Cancer Research Institute (IARC), there were 4.57 million new cancer cases including 0.48 million GC patients and 3 million cancer deaths including 0.37 million GC patients, which ranks as the third in China. 1 Metastasis is the main cause of death in patients with advanced GC. More than 40% of GC patients will have metastasis, which seriously affects the treatment, and the 5 years survival rate is less than 30% among GC patients. [1][2][3] Peritoneal dissemination may exist in early GC, which may not be curative by surgical treatment, leading to poor prognosis. 4 Peritoneal metastasis is the most common event in recurrent GC. 5,6 Preoperative peritoneal lavage fluid (PPLF) cytology is usually detected by Papanicolaou staining. 7 However, peritoneal metastasis sometimes occurs, even when cytological examination shows negative. Therefore, it is necessary to develop novel technologies or to find potential tumor markers for the diagnosis of peritoneal dissemination in GC patients, which is also helpful to improve the treatment of patients with metastatic GC. A series of biomarkers detected by reverse transcription-polymerase chain reaction (RT-PCR) in PPLF can improve peritoneal dissemination diagnosis. Common biomarkers include carcinoembryonic antigen (CEA), 8,9 carbohydrate antigen 125 (CA12-5), 10 cytokeratin 20 (CK20), 11 epithelial cell adhesion molecule-specific antigen (clone one: Ber-EP4), 12 matrix metallopeptidase 7(MMP-7), 13 Survivin, 14 Mucin-2 (MUC2), 15 interleukin-17 (IL-17), 16 fatty acid binding protein 1 (FABP1), 15 trefoil factor family 1 (TFF1), 15 and mammary serine protease inhibitor (MASPIN). 15 Most diagnostic biomarkers of peritoneal dissemination are also prognostic markers. Positivity of multiple mRNA markers could predict peritoneal recurrence and poor outcomes in GC patients with cytology negative.
Recently, methylation-specific PCR (MSP) technology has been successfully used to analyze methylation markers in liquid biopsy specimens, including PPLF samples for peritoneal metastasis diagnosis. [17][18][19] Our previous studies suggest that DNA methylation in PPLF or serum may be a potential marker for detecting peritoneal dissemination in GC patients. [20][21][22][23][24][25] Thrombospondin 1 (THBS1) is a subunit of the disulfide chain homologous trimer protein and a cohesive glycoprotein that mediates cell-to-cell and cell-to-matrix interactions. THBS1 can bind to fibrinogen, fibronectin, laminin, type V collagen, and integrin-V/-1, which has been proven to play an important role in angiogenesis, platelet aggregation, and tumorigenesis. 26 However, the role of THBS1 in tumors remains somewhat controversial. Some studies reveal that THBS1 promotes tumor cell invasion and metastasis in breast, gastric, and pancreatic cancers. [27][28][29] Another study suggests that inhibition of tumor growth by THBS1 is thought to be associated with its antiangiogenic activity. 30 Aberrant THBS1 methylation was detected in gastric cardia adenocarcinoma, melanoma, colorectal cancer, and so on, which is believed to promote tumorigenesis through its effect on angiogenesis. [31][32][33] In the present study, we explored the methylation status of THBS1 in GC and its clinicopathological significance, especially the diagnostic value of methylated THBS1 DNAs in liquid biopsy samples for peritoneal dissemination.

| Clinical tissue specimens and related information collection
The studies were performed in accordance with the Declaration of PPLF from 92 GC patients was collected in the operating room, as described previously, 21,24 clarified by centrifugation at 1500 rpm for 10 min at 4°C. DNAs from PPLF specimens were extracted for analysis of THBS1 gene methylation, and some sediments were coated on one or more glass slides and stained with Papanicolaou Besides that, antral mucosa biopsy specimens from 88 noncancer volunteers were randomly collected by gastroscopy served as controls (54 male and 34 female; mean age 52.9 years). Among these volunteers, 48 patients were diagnosed with chronic, non-atrophic gastritis, and 40 people were diagnosed with health checkup without other digestive system diseases. Paired blood samples and related information were collected from 88 non-cancer volunteers. All volunteers provided written informed consent.

| Quantitative MSP analysis
DNAs were extracted from gastric biopsies from 88 volunteers and tumor tissues from 92 GC patients, as described previously, [21][22][23][24] wherein the collected target cells were treated with 200 μg/ml pro- proportion of THBS1 methylation in the samples was calculated according to our previous reports. [20][21][22][23][24][25] In the present study, the cutoff threshold for THBS1 methylation was set to 20% based on the control normal samples and the internal quality controls provided in quantitative MSP analysis. individuals. The incidence of THBS1 methylation was significantly higher in GC tissues than that in the non-cancer control group (p < 0.0001), with no significant difference found in the incidence of THBS1 methylation between PCHNTs and controls (p>0.05). These results suggested that THBS1 methylation may be a potential marker for GC patients.

| Level of THBS1 methylation is closely related to progression in GC patients
The clinicopathological correlation analysis of THBS1 methylation in GC tissues, paired PPLF, and serum was summarized below (Table 1).
Results showed that THBS1 methylation levels in GC tissues, paired PPLF, and serum were significantly associated with tumor size, histological differentiation, lymphatic invasion, venous invasion, invasive depth, lymph node involvement, distant metastasis, and clinical

| Circulating methylated THBS1 DNAs is a novel adverse prognostic factor in GC patients
Univariate analysis showed that THBS1 methylation levels in tumor  were significantly associated with adverse prognosis in GC and worsened with the increase in THBS1 methylation degree ( Figure 3D-F).

Multivariate survival analysis revealed an independent survival
disadvantage in GC patients with THBS1 methylation in tumor tissue, PPLF, or serum specimens (p < 0.01) ( Table 3). In addition, PLCpositive (p = 0.000) or growth type (p = 0.009) parameter could be an independent influencing factor of prognosis in GC patients after surgery (Table 3). However, no significant association was estab-  Table 3).

| DISCUSS ION
The prognosis of patients with advanced GC is poor even after radical surgery. 1-3 Peritoneal dissemination is the most common type of spread, mainly caused by the seeding of free cancer cells from the primary GC. The peritoneum is also the most frequent site of recurrence in GC patients who underwent curative resection. 5,6 The examination of free tumor cells in PPLF in GC patients revealed peritoneal dissemination, predicted peritoneal recurrence and prognosis, and selected proper treatments. 7,38 Currently, cytology and molecular biology are the two most popular methods for the detection of peritoneal dissemination. [7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25] In previous studies, aberrant DNA promoter methylation of tumor-related genes has been identified in the body fluids for In this study, the detection of THBS1 methylation in GC was feasible, with highly consistent levels of THBS1 methylation detected in serum, PPLF, and tumor tissues in the same patient. In addition, the status of THBS1 methylation has a high correlation with clinicopathological parameters; that is, it can predict peritoneal dissemination and is more sensitive than peritoneal cytology.
Some studies have shown that THBS1 methylation is an early and frequent event in various tumors that can occur at each stage in the tumor and gradually increases as tumor progresses. [31][32][33] Our results suggested that THBS1 methylation is a new marker for the diagnosis of peritoneal transmission and can serve as a novel prognostic biomarker for GC patients. The current results confirmed that circulating methylated THBS1 DNAs is a potential poor prognostic factor in GC patients.
In conclusion, this was the first prospective study of the epigen- dissemination compared with peritoneal cytology. Abnormal THBS1 methylation was readily detected in preoperative serum samples in GC patients with peritoneal dissemination, and it is therefore considered a valuable biomarker for assessing the risk of peritoneal dissemination and the early diagnosis of peritoneal micrometastasis.
All in all, it is potentially significant for evaluating the high risk of peritoneal dissemination and monitoring the clinical treatment and prognosis of GC patients.

CO N FLI C T S O F I NTE R E S T
This study is not related to any potentially competing financial or other interests.

AUTH O R CO NTR I B UTI O N S
Hu XY, Ling ZN, Hong LL, Li P, and Ling ZQ contributed to the design, execution, and analysis of data. Yu QM provided clinical samples including GC tissue, serum, and PPLF samples. Hu XY, Li P, and Ling ZQ drafted the manuscript. Ling ZN and Hong LL provided some help for data analysis. All the authors were involved in the critical revision of the manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data used to support the findings of this study are available from the corresponding author upon request.