In our current study, we explored the relationship between postoperative anticoagulation and long-term survival in patients undergoing radical gastrectomy for GC. Our results shows that postoperative anticoagulation can significantly improve OS in patients with GC, but there is no significant correlation with DFS. This enlightens us that routine anticoagulation after gastrectomy is necessary.
Through multivariate survival analysis of anticoagulation status, tumor markers, blood routine, coagulation indexes, and pathological features, we confirmed that anticoagulation can significantly improve OS in patients after radical gastrectomy. Coagulation activation is strongly associated with tumor progression. The improvement of prognosis may be related to the influence of anticoagulant drugs on tumor progression. Pathological activation of the coagulation system occurs in almost all forms of cancer, particularly epithelial malignancies. There is a bidirectional interaction between coagulation activity and cancer, and the activation of the cancer-associated pathological hemostatic system promotes the progression of malignant tumors[13]. This was supported by studies showing that men with hemophilia have a lower incidence of prostate cancer[14]. Tissue factors may affect angiogenesis and the pretumorigenic inflammatory environment, thereby supporting tumor development[15]. Microthrombi in the blood protect circulating tumor cells from attack by immune cells[16]. In fact, cancer cells can activate the hemostatic system through the release of tissue factors, the expression of procoagulant proteins, the exposure of procoagulant lipids, the release of inflammatory cytokines and microparticles, and the adhesion to host vascular cells[17, 18]. Moreover, LMWH downregulates tissue factor expression and activity by modulating growth factor receptor-mediated induction of nuclear factor-κB[19]. This evidence is consistent with our findings. By comparing baseline data, we found that DFS in the anticoagulant group was significantly higher than that in the non-anticoagulant group (median: 28 vs 20 months). Although anticoagulation was not significantly associated with DFS in the multivariate survival analysis, this may be related to our relatively small sample size.
Although many studies have reported a close correlation between tumor hypercoagulability and tumor progression, the improvement of OS by postoperative anticoagulant therapy may be related to its effect on cancer-associated thrombosis. Cancer-associated thrombosis, including arterial thrombosis and VTE, significantly affects the quality of life and shortens survival. Although risk assessment tools have been shown to be effective in identifying patients at risk for VTE (including Padua prediction score, Caprini risk assessment model, IMPROVE models, the Geneva risk score and the Kucher score)[20–22], and there have been advances in perioperative management and prophylaxis measures, the risk of VTE has increased in recent years. The risk of VTE is related to the type of cancer GC is one of the cancers with the highest risk of VTE as well as individual patient differences and new anticancer therapies including antiangiogenic drugs, multi-targeted tyrosine kinase inhibitors, immunomodulatory drug combinations[8, 23, 24]. 1-year cumulative risk of pulmonary embolism and deep vein thrombosis for gastroesophageal cancer was 2.3% and 2.19%, respectively. The etiology-specific risk of pulmonary embolism and deep vein thrombosis increased by 10–30 times in patients undergoing surgery for gastroesophageal cancer, and the post-operative HR gradually decreases until it levels off, but remains higher than that in the general population[8]. Therefore, the earlier the use of anticoagulant therapy after surgery, the more effective the prevention of VTE. A large clinical study reported that GC resection exhibited the highest rates of VTE (4.1%). The diagnosis of VTE was associated with increased mortality (10.2% vs 1.7%) and prolonged hospital stay (19.5 days vs 7.5 days), as well as increased rates of readmission and admission costs[24]. Therefore, early use of anticoagulant drugs after surgery can effectively reduce VTE events and thus reduce mortality. According to ASCO VTE guideline, patients who are candidates for extended pharmacologic thromboprophylaxis after surgery may be offered prophylactic doses of LMWH (Strength of recommendation: Strong)[25]. This may explain our finding that postoperative anticoagulation prolongs OS in patients with GC. Non-anticoagulation was an independent risk factor for the long-term prognosis of GC (HR: 1.397, 95% CI: 1.018–1.918).
In addition to the anticoagulant mechanism, heparin also exerts anticancer properties through a non-anticoagulant mechanism[26]. Heparin can affect cancer cell proliferation, adhesion, angiogenesis, migration, and invasion through a variety of mechanisms, including inhibition of heparinase, P-/L-selectin, angiogenesis, and interference with the CXCL12-CXCR4 axis[12, 27]. Heparin directly induces apoptosis in cancer cells through cell cycle arrest, as well as enhances tumor cell sensitivity to chemotherapy by increasing drug uptake and reversing the downregulation of tumor suppressor genes[28, 29]. Nevertheless, the evidence for the potential survival effect of LMWH is controversial[12]. Moreover, few studies have reported the relationship between LMWH and the prognosis of GC. Our study provides evidence that postoperative anticoagulation prolongs OS in GC.
It is important to note that perioperative anticoagulation may increase the risk of bleeding. One study showed that patients with esophageal cancer who were treated with anticoagulation in the perioperative period had 22.2% postoperative bleeding incidence, compared with 5.8% in the non-anticoagulation group[30]. A clinical study of 1536 patients found that delayed massive hemorrhage after gastrectomy accounted for 0.9%, of which 3 cases died, and early postoperative hemorrhage accounted for 1.2%, and no death cases[31]. A recent study found that preoperative and postoperative use of antithrombotic drugs did not increase bleeding complications after radical gastrectomy[32]. Similarly, another study reported that preoperative use of antiplatelet and/or anticoagulants does not significantly increase the risk of major bleeding after laparoscopic GC surgery[33]. Although our study did not calculate the rate of postoperative bleeding in the anticoagulant and non-anticoagulant groups, we can conclude that perioperative use of anticoagulant drugs is relatively safe based on the above evidence. Of course, the use of anticoagulants needs to meet clinical indications. Individualized treatment with LMWH or oral anticoagulants combined with bleeding risk and VTE risk assessment is critical.
In our study we did not investigate the survival difference between prolonged perioperative anticoagulation and postoperative anticoagulation alone. One study reported that prolonged perioperative anticoagulation did not improve DFS or OS in patients with colorectal cancer undergoing surgical resection compared with in-hospital postoperative thromboprophylaxis[34]. The optimal timing of anticoagulation therapy in patients with GC needs to be further explored.