Safety, efficacy and prognosis of anticoagulant therapy for portal vein thrombosis in cirrhosis: a retrospective cohort study

Baseline characteristics of included studies

236 patients with cirrhotic PVT were initially identified. After screening, 77 patients met the the inclusion and exclusion criteria (Fig. 1), including 27 in the anticoagulant group and 50 in the non-anticoagulant group. Anticoagulants used included warfarin with the INR target level of 1.5–2.5 (n = 6), nadroparin 4100 U qd (n = 2), heparin 12,500 U qd (n = 1), rivaroxaban 20 mg qd (n = 3), rivaroxaban 10 mg qd (n = 14), edoxaban 30 mg qd (n = 1).

Fig. 1figure 1

Patients screening Flow chart

Anticoagulant therapy was started within 3 months of estimated PVT onset in 24 (69%) of 27 patients, whereas in 3 patients, anticoagulant therapy was initiated after 8, 21, and 36 months. The median follow-up time was 26 months (IQR 13–44 months) and did not differ significantly between anticoagulant and non-anticoagulant groups (18 vs 28.5 months, P = 0.071). The median time of imaging follow-up was 10 months (IQR 5.5–22.5 months) and did not differ significantly between the two groups (10 vs 10.5 months, P = 0.712). A total of 24 of 27 patients (88.9%) who received anticoagulants eventually discontinued therapy. Patients discontinuing anticoagulant therapy because of PVT recanalization (n = 10), bleeding complication (n = 4), no response after treatment for more than 6 months (n = 3), decreased treatment adherence (n = 3), worsening clinical status (n = 1) and not clearly documented (n = 3). The median duration of anticoagulant therapy was 6 (IQR 2–11) months. The baseline characteristics of the included patients are shown in Table 1.

Table 1 Baseline characteristics of included patientsRate of PVT recanalization

A Kaplan-Meier curve depicting the probability of PVT recanalization over time among patients who did and did not receive anticoagulants is depicted in Fig. 2. A total of 12/27 (44.44%) patients who received anticoagulants experienced PVT recanalization compared with 10/50 (20%) of patients who did not receive anticoagulants (log-rank P = 0.016). One patient in the anticoagulant group had complete PVT recanalization, and 11 patients had partial PVT recanalization. In the non-anticoagulant group, complete recanalization of PVT occurred in 6 patients, and partial PVT recanalization for 4 patients. Multiple cox regression analysis adjusted by the history of splenectomy, platelet count, and hemoglobin suggested that the application of anticoagulants was associated with a significantly higher rate of PVT recanalization (HR 2.672, 95% CI 1.151–6.203, P = 0.022). There was no significant difference in PVT recanalization rate among different anticoagulants including warfarin, heparin, and DOACs (Table 2).

Fig. 2figure 2

Kaplan-Meier survival curve of PVT recanalization

Table 2 Cox regression analysis of PVT recanalization

PVT progression occurred among 13/27 (7.4%) of those who received anticoagulants compared with 15/50 (30%) of those who did not (log-rank P = 0.026) (Fig. 3). The multiple cox regression analysis suggested that the application of anticoagulants was associated with a significantly lower rate of PVT progression (HR 0.221, 95% CI 0.051–0.969, P = 0.017).

Fig. 3figure 3

Kaplan-Meier survival curve of PVT progression

Safety of anticoagulant therapy

Five bleeding episodes among 4 patients (14.8%) occurred during follow-up after anticoagulant therapy. One patient had variceal bleeding and hematochezia, and other patients had upper gastrointestinal bleeding, abnormal uterine bleeding, and major variceal bleeding, respectively. All patients in the anticoagulant group stopped anticoagulant treatment after bleeding complications. One patient developed thrombocytopenia when receiving rivaroxaban 10 mg qd, and continued treatment by reducing the frequency of administration to 10 mg qod. A total of 17 bleeding events occurred in 12 patients in the non-anticoagulant group during follow-up. One patient had four variceal bleeding and two patients had once variceal bleeding and once melena. The other 5 patients developed once variceal bleeding and 4 patients had once melena. Among the 8 patients with variceal bleeding, 3 were major bleeding. There were no significant differences in the incidence of total bleeding (14.8% vs 24%, P = 0.343), major bleeding (3.7% vs 6%, P = 0.665) and variceal bleeding (3.7% vs 16%, P = 0.109) between anticoagulant group and non-anticoagulant group. There was no significant difference in the incidence of total bleeding, variceal bleeding, and major bleeding among warfarin, heparin, and DOACs (Table 3).

Table 3 Comparison of bleeding events with different anticoagulants

The significant predictor of variceal bleeding was the history of variceal bleeding. Among 34 patients with a history of variceal bleeding, 8 patients had variceal bleeding during follow-up (23.5%); Of the 40 patients without a history of variceal bleeding, 1 developed variceal bleeding during follow-up (2.5%), and the difference was statistically significant (OR = 12.0 95% CI 1.416–101.714, P = 0.023).

Prognosis

At the end of the follow-up, there were 12 patients of Child-Pugh A, 12 of Child-Pugh B, and 1 of Child-Pugh C in the anticoagulant group. And in the non-anticoagulant group, there were 12 patients of Child-Pugh A, 24 of Child-Pugh B, and 11 of Child-Pugh C. The classification of liver function has a significant difference between the anticoagulant group and the non-anticoagulant group (P = 0.030) after follow-up. But there was no significant difference in the 2-year survival rate between the two groups (P = 1.000). One patient in the anticoagulant group died of liver failure 8 months after the diagnosis of PVT. In the non-anticoagulant group, one patient died of abdominal infection 9 months after the diagnosis, and one patient died of liver failure 24 months after the diagnosis. None of these deaths were related to bleeding complications.

留言 (0)

沒有登入
gif