Skip to main content
NCBI home page
Clinical and Applied Thrombosis/Hemostasis logoLink to Clinical and Applied Thrombosis/Hemostasis
. 2023 Jul 18;29:10760296231188718. doi: 10.1177/10760296231188718

Anticoagulation Therapy for Splanchnic Vein Thrombosis Associated With Acute Pancreatitis: A Systematic Review and Meta-Analysis

Yuhang Yin 1,2,*, Le Wang 1,2,*, Fangbo Gao 1,3,*, Lei Liu 4,✉, Xingshun Qi 1,2,3,✉
PMCID: PMC10357047  PMID: 37461391

Abstract

Splanchnic vein thrombosis (SVT) is not rare in patients with acute pancreatitis. It remains unclear about whether anticoagulation should be given for acute pancreatitis-associated SVT. The PubMed, EMBASE, and Cochrane Library databases were searched. Rates of SVT recanalization, any bleeding, death, intestinal ischemia, portal cavernoma, and gastroesophageal varices were pooled and compared between patients with acute pancreatitis-associated SVT who received and did not receive therapeutic anticoagulation. Pooled rates and risk ratios (RRs) with 95% confidence intervals (CIs) were calculated. Heterogeneity among studies was evaluated. Overall, 16 studies including 698 patients with acute pancreatitis-associated SVT were eligible. After therapeutic anticoagulation, the pooled rates of SVT recanalization, any bleeding, death, intestinal ischemia, portal cavernoma, and gastroesophageal varices were 44.3% (95%CI = 32.3%-56.6%), 10.7% (95%CI = 4.9%-18.5%), 13.3% (95%CI = 6.9%-21.4%), 16.8% (95%CI = 6.9%-29.9%), 21.2% (95%CI = 7.5%-39.5%), and 29.1% (95%CI = 16.1%-44.1%), respectively. Anticoagulation therapy significantly increased the rate of SVT recanalization (RR = 1.69; 95%CI = 1.29-2.19; P < .01), and marginally increased the risk of bleeding (RR = 1.98; 95%CI = 0.93-4.22; P = .07). The rates of death (RR = 1.42; 95%CI = 0.62-3.25; P = .40), intestinal ischemia (RR = 2.55; 95%CI = 0.23-28.16; P = .45), portal cavernoma (RR = 0.51; 95%CI = 0.21-1.22; P = .13), and gastroesophageal varices (RR = 0.71; 95%CI = 0.38-1.32; P = .28) were not significantly different between patients who received and did not receive anticoagulation therapy. Heterogeneity was statistically significant in the meta-analysis of intestinal ischemia, but not in those of SVT recanalization, any bleeding, death, portal cavernoma, or gastroesophageal varices. Anticoagulation may be effective for recanalization of acute pancreatitis-associated SVT, but cannot improve the survival. Randomized controlled trials are warranted to further investigate the clinical significance of anticoagulation therapy in such patients.

Keywords: acute pancreatitis, splanchnic vein, thrombosis, anticoagulation, survival

Graphical Abstract.

Graphical Abstract

Open in a new tab

This is a visual representation of the abstract.

Introduction

Splanchnic vein thrombosis (SVT) is a serious vascular complication of acute pancreatitis.1‐3 The incidence of acute pancreatitis-associated SVT varies from 16.6% to 22.6% among studies, depending on the severity of acute pancreatitis and imaging techniques used.4‐6 Acute pancreatitis-associated SVT involves the portal vein (PV), the splenic vein (SV), and the superior mesenteric vein (SMV) in their combination or alone.3,7 Development of SVT in patients with acute pancreatitis is mainly related to local inflammation and compression by pancreatic enlargement or pseudocyst.8,9 It usually occurs within 1 to 2 weeks after the onset of moderately severe and severe acute pancreatitis, which may prolong the disease course and increase the mortality. 7 SVT may be asymptomatic in most cases, but can cause intestinal ischemia secondary to SMV obstruction and develop portal cavernoma and gastroesophageal variceal bleeding attributed to regional portal hypertension.10‐12

Currently, the optimal management of acute pancreatitis-associated SVT is still controversial. Therapeutic anticoagulation has been recommended when SVT involves mesenteric vein and causes abdominal symptoms related to intestinal ischemia. 12 This recommendation is primarily based on negative impact of acute symptomatic SMV thrombosis on the patients’outcomes. 13 Previous meta-analyses have explored the role of therapeutic anticoagulation in acute pancreatitis-associated SVT,14‐17 but their conclusions remain controversial. Besides, a previous meta-analysis had some obvious drawbacks in term of study design. 14 First, its selection criteria were problematic, because an original study regarding prophylactic anticoagulation was incorrectly included. 18 Second, its outcome assessment was inaccurate, because an eligible study was omitted from the meta-analysis of bleeding complications. 19 For these reasons, we have performed an updated meta-analysis using more strict inclusion criteria and accurately extracting the relevant data to further compare the rates of SVT recanalization, any bleeding, death, intestinal ischemia, portal cavernoma, gastroesophageal varices, major bleeding, and gastrointestinal bleeding (GIB) in patients with acute pancreatitis-associated SVT who received therapeutic anticoagulation and did not receive anticoagulation.

Methods

The systematic review with meta-analysis was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement as shown in the Supplemental Material.

Registration

The study was registered in PROSPERO with registration no. CRD 42022171698.

Literature Source

All published papers were searched via the PubMed, EMBASE, and Cochrane Library databases. The last search was conducted on July 14, 2022. Search items were as follows: (pancreatitis [all fields]) AND ((portal vein [all fields]) OR (portal venous [all fields]) OR (mesenteric vein [all fields]) OR (mesenteric venous [all fields]) OR (splanchnic vein [all fields]) OR (splanchnic venous [all fields]) OR (portosplenomesenteric vein [all fields]) OR (portosplenomesenteric venous [all fields])) AND ((thrombosis [all fields]) OR (obstruction [all fields]) OR (occlusion [all fields])) AND ((anticoagulant [all fields]) OR (anticoagulation [all fields])).

Study Selection

All eligible studies should have compared the outcomes of patients with acute pancreatitis-associated SVT who received and did not receive anticoagulants. Exclusion criteria were as follows: (1) duplicates; (2) case reports; (3) letters, comments, or editorials; (4) reviews and/or meta-analyses; (5) guidelines, notes, or reports; (6) experimental or animal studies; (7) patients without acute pancreatitis and/or SVT; (8) anticoagulation was not given in any patient; and (9) any outcome of interest, including SVT recanalization, any bleeding, death, portal cavernoma, gastroesophageal varices, intestinal ischemia, major bleeding, and GIB, was not reported. Neither publication language nor publication year was restricted.

Data Extraction

The following data were extracted from each study: first author, publication year, country, study design, type of publication, enrollment period, number of patients with acute pancreatitis-associated SVT, severity of acute pancreatitis, extension of SVT, number of patients who received and did not receive anticoagulation, type and dosage of anticoagulants, duration of anticoagulation, and number of patients who developed SVT recanalization, any bleeding, death, intestinal ischemia, portal cavernoma, gastroesophageal varices, major bleeding, and GIB.

Outcomes

The primary outcomes included SVT recanalization and any bleeding. The secondary outcomes included death, intestinal ischemia, portal cavernoma, gastroesophageal varices, major bleeding, and GIB.

Definitions

According to the International Society on Thrombosis and Haemostasis (ISTH) definition, major bleeding often refers to a fatal event, including bleeding from major organ with obvious symptoms and bleeding at the surgical site requiring reoperation or leading to hospitalization. 20 Based on the ISTH criteria and the definition proposed by every individual study included, we defined major bleeding as arterial bleeding caused by parenchymal organ and intra-abdominal bleeding from rupture of pseudocysts.

GIB mainly included bleeding due to peptic ulcer and gastroesophageal varices secondary to regional portal hypertension. 21

Study Quality

The quality of cohort studies was evaluated by the Newcastle-Ottawa Scale in terms of 3 major parts: (1) selection (score 0-4); (2) comparability (score 0-2); and (3) outcome (score 0-3). A score of 0-3, 4-6, and 7-9 was deemed as low, moderate, and high quality, respectively.

Statistical Analyses

Meta-analyses were conducted by using a random-effects model in the Review Manager 5.3 (Cochrane Collaboration, Nordic Cochrane Centre, Copenhagen), StatsDirect 2.8.0 (Stats-Direct Ltd, Sale, Cheshire, UK), and STATA 12.0 (Stata Corp, College Station, TX). Pooled rates and risk ratios (RRs) with their 95% confidence intervals (CIs) were calculated. Cochrane Q test and I² statistics were used to assess heterogeneity among studies, and P < .1 or I² > 50% was considered as statistically significant heterogeneity. Subgroup, meta-regression, and sensitivity analyses were employed to explore the sources of heterogeneity. Both subgroup and meta-regression analyses were performed in terms of publication year (before vs after 2015), study design (prospective vs retrospective), region (America vs Non-America), type of publication (full texts vs abstracts), study quality (moderate vs high), indication of anticoagulation (specific vs not specific to intestinal ischemia), and sample size (≤40 vs >40). Interaction between subgroups was explored. P < .1 was considered as statistically significant interaction. Leave-one-out sensitivity analyses were performed by removing studies one by one from a meta-analysis. Publication bias was evaluated with the Egger test, and P < .1 was considered as statistically significant publication bias.

Results

Study Characteristics

Initially, 434 papers were identified. A total of 16 studies including 698 patients with acute pancreatitis-associated SVT were finally eligible (Figure 1). Characteristics of included studies are shown in Table 1. There were 14 retrospective cohort studies1,10,19,22‐32 and 2 prospective cohort studies.33,34 The sample size ranged from 7 to 128. Nine studies were published as full texts1,10,19,22,26‐28,30,33 and 7 as abstracts.23‐25,29,31,32,34 Studies were published between 2008 and 2022. Six studies were performed in Europe,19,22,26‐28,30 2 in Asia,25,33 and 8 in North America.1,10,23,24,29,31,32,34 Eleven studies19,23‐25,27‐32,34 were of moderate quality and 51,10,22,26,33 of high quality (Supplemental Table 1).

Figure 1.

Figure 1.

Open in a new tab

Flow chart of study selection.

Abbreviation: SVT, splanchnic vein thrombosis.

Table 1.

Characteristics of Studies Regarding Pancreatitis and SVT.

First author (year) Country Study design Type of publication Enrollment period No. Pts. pancreatitis and SVT Type and severity of pancreatitis Extension of SVT No. Pts. AC/Non-AC Type and dosage of anticoagulants Duration of anticoagulation (months) NOS score
K (2022) UK Retrospective cohort Full-text 2018-2021 109 MAP (n = 11)
MSAP (n = 60)
SAP (n = 38)		 SV (n = 46)
SMV/PV (n = 36)
SMV + PV + SV (n = 27)		 74/35 LMWH; Apixaban Range: 0.2-6 6
Pagliari (2020) Italy Retrospective cohort Full-text 2015.12-2018.12 27 MAP (n = 4)
MSAP (n = 14)
SAP (n = 9)		 PV (n = 1)
SV (n = 9)
PV + SMV (n = 1)
PV + SV (n = 2)
SV + SMV (n = 10)
PV + SMV + SV (n = 4)		 16/11 LMWH: 100 UI/kg, bid; Warfarin (NA); Fondaparinux: 7.5 mg/day; DOACs: 5 mg, bid; Loading dose: NA Mean ± SD: 5.2 ± 2.2 8
Clark (2020) USA Retrospective cohort Abstract 2016.06-2019.12 35 MSAP and SAP (NA) SV (n = 13)
PV + SMV (n = 22)		 32/3 NA NA 4
Junare (2020) India Prospective cohort Full-text 2018.01-2018.12 24 AP (n = 24) SV (n = 11)
PV + SV (n = 4)
PV + SMV + SV (n = 9)		 12/12 UFH; Warfarin NA 7
Garret (2018) France Retrospective cohort Full-text 2012.01-2015.12 76 AP (n = 76) SV (n = 62)
Unknown (n = 14)		 39/37 NA Mean ± SD: 5.68 ± 1.02 6
Wesley (2017) USA Retrospective cohort Abstract 2007.01-2017.01 128 AP (n = 128) PV (n = 48)
SV (n = 57)
PV + SV (n = 23)		 57/71 NA NA 5
Wang (2017) China Retrospective cohort Abstract 2013-2014 7 AP (n = 7) PV (n = 4)
SV (n = 2)
SMV (n = 1)		 1/6 NA NA 5
Toqué (2015) France Retrospective cohort Full-text 2007.01-2012.12 19 AP (n = 19) PV (n = 6)
SV (n = 7)
SV + SMV (n = 4)
PV + SMV + SV (n = 2)		 15/4 NA <6 (n = 10); >6 (n = 5) 8
Yang (2015) USA Retrospective cohort Abstract 2004-2013 21 AP (n = 21) PV (n = 3)
SV (n = 7)
SMV (n = 3)
PV + SMV (n = 2)
PV + SV (n = 2)
SV + SMV (n = 3)
PV + SMV + SV (n = 1)		 12/9 UFH; Warfarin; Direct thrombin inhibitor NA 4
Easler (2014) USA Retrospective cohort Full-text 2003.06-2010.04 22 SAP (n = 22) PV (n = 1)
SV (n = 13)
SMV (n = 1)
PV + SMV (n = 1)
PV + SV (n = 2)
PV + SMV + SV (n = 4)		 4/18 NA <8 (n = 3); >8 (n = 1) 7
Hall (2013) UK Retrospective cohort Full-text 1997.12-2010.12 11 AP (n = 11) PV + SMV (n = 1)
PV + SV (n = 3)
SV + SMV (n = 2)
Unknown (n = 5)		 7/4 LMWH ± Warfarin; UFH ± Warfarin; Warfarin alone Range: 2-lifelong 5
Harris (2013) USA Retrospective cohort Full-text 1996.01-2006.12 45 MAP (n = 19)
SAP (n = 26)		 PV (n = 7)
SV (n = 17)
SMV (n = 4)
PV + SMV (n = 4)
PV + SV (n = 4)
SV + SMV (n = 4)
PV + SMV + SV (n = 5)		 17/28 LMWH: 1 mg/kg/12 h; Loading dose: UFH, 80 U/kg, followed by 18/kg/hour; Warfarin (NA) Range: 3-12 7
Muddana (2012) USA Prospective cohort Abstract 2003-2010 41 SAP (n = 41) PV + SMV + SV (n = 8)
Unknown (n = 33)		 7/34 NA NA 5
Gonzelez (2011) UK Retrospective cohort Full-text 2008.01-2009.12 20 AP (n = 20) PV (n = 5)
SV (n = 8)
SMV (n = 1)
PV + SV (n = 4)
SV + SMV (n = 1)
PV + SMV + SV (n = 1)		 4/16 LMWH: 1 mg/kg; Warfarin (NA)
Loading dose: NA		 NA 6
Pribramska (2009) USA Retrospective cohort Abstract 1989-2007 50 AP (n = 50) PV (n = 3)
SV (n = 29)
SMV (n = 4)
PV + SV (n = 4)
PV + SMV (n = 2)
PV + SMV + SV (n = 4)		 8/42 NA NA 5
Harris (2008) USA Retrospective cohort Abstract NA 63 SAP (n = 63) PV (n = 9)
SV (n = 25)
SMV (n = 5)
PV + SMV (n = 5)
PV + SV (n = 7)
SV + SMV (n = 5)
PV + SMV + SV (n = 7)		 25/38 UFH 6 5
Open in a new tab

Abbreviations: AC, anticoagulation; Non-AC, non-anticoagulation; AP, acute pancreatitis; Pts, patients; SVT, splanchnic vein thrombosis; PV, portal vein; SMV, superior mesenteric vein; SV, splenic vein; NA, not available; MAP, mild acute pancreatitis; MSAP, moderate severe acute pancreatitis; SAP, severe acute pancreatitis; UFH, unfractionated heparin; LMWH, low molecular weight heparin; DOACs, direct oral anticoagulants; NOS, Newcastle–Ottawa Scale.

Patient Characteristics

Seven studies,1,10,22,23,30,32,34 including 342 patients, reported the severity of acute pancreatitis. Among them, 50 (14.6%), 93 (27.2%), and 199 (58.2%) had mild, moderately severe, and severe acute pancreatitis, respectively. All of the 16 studies reported the extension of SVT (Table 1). Isolated SV (306/698, 43.84%) was the most frequently involved vessel, followed by isolated PV (87/698, 12.46%), a combination of PV, SV, and SMV (82/698, 11.75%), a combination of PV and SV (55/698, 7.89%), a combination of PV and SMV (38/698, 5.44%), a combination of SV and SMV (29/698, 4.15%), and isolated SMV (19/698, 2.72%).

Anticoagulation

Only one study clearly reported the indications of anticoagulation, including symptoms of bowel ischemia and/or hepatic decompensation defined as new onset of jaundice, ascites, and hepatic encephalopathy. 33 Another study clearly reported the timing of anticoagulation, and showed that anticoagulation was given on the day of diagnosis of acute SVT. 27 Seven studies reported the type and dosage of anticoagulants.10,22,27,28,31‐33 Type of anticoagulants included low molecular weight heparin (LMWH) alone (n = 5), warfarin alone (n = 6), unfractionated heparin (UFH) alone (n = 5), direct thrombin inhibitor alone (n = 1), fondaparinux followed by direct oral anticoagulants (DOACs) (n = 1), LMWH followed by warfarin (n = 1), and UFH followed by warfarin (n = 1). Dosage of LMWH was 100 UI/kg twice a day (n = 2), that of UFH 80 U/kg followed by 18 U/kg per hour (n = 1), and that of DOACs (apixaban) 5 mg twice a day (n = 1) (Table 1).

Primary Outcomes

Recanalization

Twelve studies reported the rate of SVT recanalization, and the pooled rate was 44.3% (95%CI = 32.3%-56.6%) in patients who received anticoagulation and 20.6% (95%CI = 14.0%-28.2%) in those who did not receive anticoagulation. The rate of SVT recanalization was significantly higher in patients who received anticoagulation than in those who did not receive anticoagulation (RR = 1.69; 95%CI = 1.29-2.19; P < .01) (Figure 2). Heterogeneity was not statistically significant (I² = 0%; P = .56). Thus, meta-regression and sensitivity analyses were not performed. The benefit of anticoagulation on SVT recanalization remained statistically significant in subgroup analyses of moderate-quality studies (43.1% vs 20.3%; RR = 1.72; P < .01), retrospective studies (44.8% vs 20.2%; RR = 1.76; P < .01), studies published as full-texts (51.9% vs 23.1%; RR = 1.86; P < .01) and abstracts (33.8% vs 18.3%; RR = 1.80; P = .03), studies where the indication of anticoagulation was not specific to intestinal ischemia (42.4% vs 19.1%; RR = 1.74; P < .01), and studies with a sample size of >40 (35.3% vs 19.7%; RR = 1.67; P < .01) and ≤40 (55.7% vs 23.0%; RR = 2.25; P < .01), but not those of high-quality studies (46.5% vs 21.3%; RR = 1.67; P = .10), prospective studies (42.6% vs 24.6%; RR = 1.44; P = .35), or studies where the indication of anticoagulation was specific to intestinal ischemia (50.0% vs 41.6%; RR = 1.20; P = .68). The interaction was not statistically significant among all subgroups (Supplemental Table 2). Publication bias was statistically significant (P = .07).

Figure 2.

Figure 2.

Open in a new tab

Forest plots comparing the rate of SVT recanalization between anticoagulation and non-anticoagulation groups.

Abbreviations: SVT, splanchnic vein thrombosis; AC, anticoagulation; Non-AC, non-anticoagulation.

Any bleeding

Twelve studies reported the rate of any bleeding, and the pooled rate was 10.7% (95%CI = 4.9%-18.5%) in patients who received anticoagulation and 5.7% (95%CI = 2.0%-11.1%) in those who did not receive anticoagulation. The rate of any bleeding was not significantly different between the 2 groups (RR = 1.98; 95%CI = 0.93-4.22; P = .07) (Figure 3). Heterogeneity was not statistically significant (I² = 26.2%; P = .22). Thus, meta-regression and sensitivity analyses were not performed. The effect of anticoagulation on any bleeding remained statistically significant in subgroup analyses of moderate-quality studies (9.1% vs 3.1%; RR = 2.46; P < .01) and studies with a sample size of >40 (15.1% vs 7.1%; RR = 2.46; P < .01), but not those of high-quality studies (15.1% vs 7.6%; RR = 1.66; P = .56), studies published as full-texts (12.9% vs 6.1%; RR = 1.76; P = .29) and abstracts (6.6% vs 2.2%; RR = 2.49; P = .08), retrospective (10.3% vs 5.1%; RR = 1.80; P = .16) and prospective studies (13.7% vs 1.0%; RR = 7.00; P = .18), studies with a sample size of ≤40 (11% vs 3.1%; RR = 1.50; P = .71), or studies where the indication of anticoagulation was specific to intestinal ischemia (33.3% vs 0%; RR = 7.00; P = .18) and not specific to intestinal ischemia (9.7% vs 4.4%; RR = 1.80; P = .16). The interaction was not statistically significant among all subgroups (Supplemental Table 3). Publication bias was not statistically significant (P = .69).

Figure 3.

Figure 3.

Open in a new tab

Forest plots comparing the rate of any bleeding between anticoagulation and non-anticoagulation groups.

Abbreviations: AC, anticoagulation; Non-AC, non-anticoagulation.

Secondary Outcomes

Death

Eight studies reported the mortality, and the pooled rate was 13.3% (95%CI = 6.9%-21.4%) in patients who received anticoagulation and 8.0% (95%CI = 2.8%-15.5%) in those who did not receive anticoagulation. The mortality was not significantly different between the 2 groups (RR = 1.42; 95%CI = 0.62-3.25; P = .40) (Figure 4). Heterogeneity was not statistically significant (I² = 33.9%; P = .18). Thus, meta-regression and sensitivity analyses were not performed. The effect of anticoagulation on survival remained statistically significant in subgroup analyses of studies published as abstracts (20.3% vs 3.5%; RR = 3.37; P = .02), but not those of studies published as full-texts (12.4% vs 13.1%; RR = 0.91; P = .80), moderate-quality (17.1% vs 9.0%; RR = 1.48; P = .47) and high-quality studies (7.6% vs 7.0%; RR = 1.26; P = .81), or studies where the indication of anticoagulation was specific to intestinal ischemia (16.6% vs 8.3%; RR = 0.50; P = .55) and not specific to intestinal ischemia (13.5% vs 7.0%; RR = 1.62; P = .30). The interaction was statistically significant among the subgroups according to the type of publication (Supplemental Table 4). Publication bias was not statistically significant (P = .85).

Figure 4.

Figure 4.

Open in a new tab

Forest plots comparing the mortality between anticoagulation and non-anticoagulation groups.

Abbreviations: AC, Anticoagulation; Non-AC, Non-anticoagulation.

Intestinal ischemia

Three studies reported the rate of intestinal ischemia, and the pooled rate was 16.8% (95%CI = 6.9%-29.9%) in patients who received anticoagulation and 4.0% (95%CI = 0.1%-17.1%) in those who did not receive anticoagulation. The rate of intestinal ischemia was not significantly different between the 2 groups (RR = 2.55; 95%CI = 0.23-28.16; P = .45) (Supplemental Figure 1). Heterogeneity was statistically significant (I² = 50.4%; P = .13). Meta-regression analyses were not performed due to a small number of studies included. Sensitivity analyses indicated that the heterogeneity might be attributed to the study by Hall et al 27 (Supplemental Table 5). The effect of anticoagulation on intestinal ischemia remained statistically significant in subgroup analyses of high-quality studies (22.4% vs 1.2%; RR = 8.54; P = .04), but not those of studies with a sample size of ≤40 (16.5% vs 10.0%; RR = 1.42; P = .85) or studies where the indication of anticoagulation was specific to intestinal ischemia (8.2% vs 7.2%; RR = 1.42; P = .55). The interaction was statistically significant among the subgroups according to the study quality (Supplemental Table 6). Publication bias was not statistically significant (P = .58).

Portal cavernoma

Four studies reported the incidence of portal cavernoma, and the pooled incidence was 21.2% (95%CI = 7.5%-39.5%) in patients who received anticoagulation and 34.2% (95%CI = 9.8%-64.2%) in those who did not receive anticoagulation. The incidence of portal cavernoma was not significantly different between the 2 groups (RR = 0.51; 95%CI = 0.21-1.22; P = .13) (Supplemental Figure 2). Heterogeneity was not statistically significant (I² = 33.7%; P = .21). Thus, meta-regression and sensitivity analyses were not performed. The effect of anticoagulation on portal cavernoma remained statistically significant in subgroup analyses of moderate-quality studies (11.3% vs 47.6%; RR = 0.33; P < .01) and studies with a sample size of >40 (23.08% vs 67.57%; RR = 0.34; P < .01), but not those of high-quality studies (9.8% vs 2.04%; RR = 1.18; P = .13) or studies with a sample size of ≤40 (6.7% vs 24.1%; RR = 0.80; P = .75). The interaction was not statistically significant among all subgroups (Supplemental Table 7). Publication bias was not statistically significant (P = .48).

Gastroesophageal varices

Three studies reported the incidence of gastroesophageal varices, and the pooled incidence was 29.1% (95%CI = 16.1%-44.1%) in patients who received anticoagulation and 33.4% (95%CI = 10.3%-61.9%) in those who did not receive anticoagulation. The incidence of gastroesophageal varices was not significantly different between the 2 groups (RR = 0.71; 95%CI = 0.38-1.32; P = .28) (Supplemental Figure 3). Heterogeneity was not statistically significant (I² = 0%; P = .80). Thus, meta-regression and sensitivity analyses were not performed. The effect of anticoagulation on gastroesophageal varices was not statistically significant in any subgroup analysis according to the study design, study quality, sample size, and indication of anticoagulation. The interaction was not statistically significant among all subgroups (Supplemental Table 8). Publication bias was not statistically significant (P = .14).

Major bleeding

Three studies reported the rate of major bleeding, and the pooled rate was 7.0% (95%CI = 1.1%-17.6%) in patients who received anticoagulation and 3.2% (95%CI = 0.4%-8.4%) in those who did not receive anticoagulation. The rate of major bleeding was not significantly different between the 2 groups (RR = 2.87; 95%CI = 1.00-8.20; P = .05) (Supplemental Figure 4). Heterogeneity was not statistically significant (I² = 0%; P = .91). Thus, meta-regression and sensitivity analyses were not performed. The effect of anticoagulation on major bleeding remained statistically significant in subgroup analyses of moderate-quality studies (9.5% vs 3.1%; RR = 2.80; P = .05) and those with a sample size of >40 (9.5% vs 3.1%; RR = 2.80; P = .05). The interaction was not statistically significant among all subgroups (Supplemental Table 9). Publication bias could not be evaluated because only a small number of studies were included and one of them reported an incidence of 0%.

GIB

Three studies reported the rate of GIB, and the pooled rate was 4.9% (95%CI = 1.4%-10.3%) in patients who received anticoagulation and 3.9% (95%CI = 0.4%-10.6%) in those who did not receive anticoagulation. The rate of GIB was not significantly different between the 2 groups (RR = 1.56; 95%CI = 0.49-5.02; P = .45) (Supplemental Figure 5). Heterogeneity was not statistically significant (I² = 0.00%; P = .43). Thus, meta-regression and sensitivity analyses were not performed. The effect of anticoagulation on GIB was not statistically significant in any subgroup analysis according to the study quality, sample size, and indication of anticoagulation. The interaction was not statistically significant among all subgroups (Supplemental Table 10). Publication bias was not statistically significant (P = .58).

Discussion

We observed a benefit of anticoagulation on the improvement of SVT recanalization in patients with acute pancreatitis-associated SVT. This finding can be explained by the pathogenesis of acute pancreatitis-associated SVT: (1) pancreatic necrosis and local inflammatory infiltration can damage blood vessels located in the peripancreatic space, resulting in vascular endothelial injury;7,11,35 (2) tissue factors from damaged pancreas are released into the blood, ultimately triggering coagulation cascade;36‐38 (3) the levels of inflammatory mediators are elevated, thereby inducing hypercoagulable state, which lead to the deposition of platelets and fibrinogen, thereby accelerating the formation of thrombosis.9,39,40 Accordingly, anticoagulation therapy can improve systemic hypercoagulability and enhance anticoagulant properties of endothelium,41,42 thereby achieving PV recanalization. On the other hand, it has been reported that anticoagulation therapy may improve the severity of acute pancreatitis and decrease the morbidity and mortality.18,43 Thus, SVT recanalization may also be attributed to the improvement of disease severity after anticoagulation.

To date, the only available practice guidance recommends that the indication for anticoagulation is the extension of SVT into mesenteric vein, 12 which may cause fatal intestinal ischemia and infarction. 44 Among the SVT patients included in our meta-analysis, 22.24% (131/589) had involvement of mesenteric vein. Unfortunately, the number of patients with mesenteric vein thrombosis who received anticoagulation therapy was unavailable. Thus, the impact of anticoagulation on SVT recanalization in such patients remains to be elucidated.

Our meta-analysis showed that anticoagulation marginally increased the risk of any bleeding, which was inconsistent with the findings of 2 previous meta-analyses.14,15 However, it should be noticeable that the relevant data regarding effect of anticoagulants on bleeding events were incomplete in the 2 previous meta-analyses. As known, the most common source of bleeding secondary to pancreatitis-associated SVT should be gastric varices, which are caused by shunting of blood within obstructed SV secondary to compression of pancreatic tails to low-pressure vessels, such as short gastric, gastric coronary, and gastroepiploic venous system.45,46 Venous recanalization by anticoagulants can enhance blood supply to the liver, thereby decreasing PV pressure and risk of bleeding from variceal rupture. 21 Similarly, it seems that anticoagulation therapy also reduces the risk of gastroesophageal variceal bleeding to some extent in cirrhotic patients with SVT, because it is potentially effective for further reduction of PV pressure after SVT recanalization. 21

Another finding of our meta-analysis was that anticoagulation therapy did not significantly decrease the risk of death. Pancreatic necrosis and multiple organ failure related to systemic inflammation are the most common causes for death in patients with acute pancreatitis.10,19,28 By comparison, few patients with acute pancreatitis die of SVT-related complications, such as intestinal ischemia or infarction and left-sided portal hypertension. In details, several characteristics of the patients with acute pancreatitis-associated SVT included in our meta-analysis should be considered to explain this unexpected phenomenon that anticoagulation failed to improve the survival. First, only a minority of deaths were due to intestinal ischemia/necrosis (∼10%). Second, a majority of SVT were located at SV alone without involvement of mesenteric vein (∼66%), which could hardly cause intestinal ischemia/necrosis. Third, the information regarding presence of intestinal ischemia or infarction before anticoagulation was unclear in most of cases. Indication of anticoagulation (presence vs absence of intestinal ischemia) may influence the judgement of patients’ outcomes (improvement vs aggravation of intestinal ischemia). Taken together, based on the currently available data, the benefits of anticoagulation on SVT recanalization have not been translated into its effects on the improvement of patients’survival. 47

Our meta-analysis has several advantages as follows. First, only 5 individual studies were included in one previous meta-analysis 15 and only 7 in another 2 previous meta-analyses.16,17 By comparison, our meta-analysis further expanded the search strategy and included 8 additional studies,23‐25,27,29,31,32,34 which made our statistical results more comprehensive and representative. Second, the selection of eligible studies in our meta-analysis was more rigorous and reasonable. In details, all eligible participants should be diagnosed with acute pancreatitis-associated SVT, and therapeutic anticoagulation was selectively given. However, in a previous meta-analysis, 14 one individual study 19 regarding the use of prophylactic anticoagulation did not meet the pre-specified inclusion criteria, but was mistakenly included. Third, multiple outcomes of interests were analyzed, including SVT recanalization, any bleeding, death, intestinal ischemia, portal cavernoma, gastroesophageal varices, major bleeding, and GIB, to more comprehensively clarify the role of anticoagulation therapy in patients with acute pancreatitis-associated SVT. Fourth, whether the indication of anticoagulation influences the patients’outcomes has been elucidated in our subgroup analyses. However, this issue has not been explored yet in previous meta-analyses.14‐17 Fifth, only a random-effects model was employed. Accordingly, our statistical results should be more conservative.

Our study has several limitations. First, no randomized controlled trials have been performed yet, suggesting that the quality of current evidence is relatively poor. Second, 7 of the 15 included studies were published as conference abstracts, where the quality of data provided is often controversial. 48 However, our meta-analyses should be more comprehensive by including all available data regardless of type of publication. The type, dosage, and duration of anticoagulants were insufficiently provided in 7 studies published only as abstracts, compromising further subgroup analyses and interpretation of secondary outcomes. Third, the severity of acute pancreatitis, degree of occlusion, intensity and duration of anticoagulation, and recurrent thrombosis were not completely reported, which may affect our judgment on the SVT recanalization and patients’outcomes. Fourth, the information regarding a combination of other antithrombotic treatments, which may strengthen the possibility of SVT recanalization, were unknown. Fifth, the presence, severity, and intervention of portal hypertension and its complications were related to the risk of bleeding, but the information regarding portal hypertension, esophageal varices, and prophylaxis and treatment of portal hypertension was not reported in any included study. In addition, the source of gastro-intestinal bleeding was not determined in any included study, therefore its association with anticoagulation could not be explored.

Conclusion

Our meta-analysis supported a potential benefit of anticoagulation therapy on SVT recanalization in patients with acute pancreatitis-associated SVT, but with a bleeding tendency. In future, the effect of anticoagulation therapy on the outcomes of patients with acute pancreatitis-associated SVT needs to be elucidated in high-quality prospective studies.

Supplemental Material

sj-docx-1-cat-10.1177_10760296231188718 - Supplemental material for Anticoagulation Therapy for Splanchnic Vein Thrombosis Associated With Acute Pancreatitis: A Systematic Review and Meta-Analysis
Click here for additional data file. (309.1KB, docx)

Supplemental material, sj-docx-1-cat-10.1177_10760296231188718 for Anticoagulation Therapy for Splanchnic Vein Thrombosis Associated With Acute Pancreatitis: A Systematic Review and Meta-Analysis by Yuhang Yin, Le Wang, Fangbo Gao, Lei Liu and Xingshun Qi in Clinical and Applied Thrombosis/Hemostasis

Acknowledgments

We are indebted to Dr Walter Ageno for his valuable comments on the language and content of our manuscript.

Abbreviations

SVT

splanchnic vein thrombosis

PV

portal vein

SV

splenic vein

SMV

superior mesenteric vein

NOS

Newcastle-Ottawa Scale

RRs

risk ratios

CIs

confidence intervals.

Footnotes

Author Contributions: Conceptualization: Xingshun Qi; Methodology: Yuhang Yin, Le Wang, Fangbo Gao, Lei Liu, and Xingshun Qi; Formal analysis: Yuhang Yin, Le Wang, Fangbo Gao, and Xingshun Qi; Data curation: Yuhang Yin, Le Wang, and Fangbo Gao; Data interpretation: Yuhang Yin, Le Wang, Fangbo Gao, Lei Liu, and Xingshun Qi; Writing-original draft: Yuhang Yin, Le Wang, Fangbo Gao, Lei Liu, and Xingshun Qi; Writing-review and editing: Yuhang Yin, Le Wang, Fangbo Gao, Lei Liu, and Xingshun Qi; Project administration: Xingshun Qi and Lei Liu. All authors have made an intellectual contribution to the manuscript and approved the submission.

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding: The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Program for Youth and middle aged Scientific and Technical Innovation Excellent Talents in Shenyang (grant number RC210011); Outstanding Youth Foundation of Liaoning Province (grant number 2022-YQ-07).

Supplemental Material: Supplemental material for this article is available online.

References

  • 1.Easler J, Muddana V, Furlan A, et al. Portosplenomesenteric venous thrombosis in patients with acute pancreatitis is associated with pancreatic necrosis and usually has a benign course. Clin Gastroenterol Hepatol. 2014;12(5):854‐862. [DOI] [PubMed] [Google Scholar]
  • 2.Mederos MA, Reber HA, Girgis MD. Acute pancreatitis: a review. Jama. 2021;325(4):382‐390. [DOI] [PubMed] [Google Scholar]
  • 3.Jiang W, Zhou J, Ke L, et al. Splanchnic vein thrombosis in necrotizing acute pancreatitis: detection by computed tomographic venography. World J Gastroenterol. 2014;20(44):16698‐16701. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Xu W, Qi X, Chen J, et al. Prevalence of splanchnic vein thrombosis in pancreatitis: a systematic review and meta-analysis of observational studies. Gastroenterol Res Pract. 2015;2015:245460. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Norton W, Lazaraviciute G, Ramsay G, et al. Current practice of anticoagulant in the treatment of splanchnic vein thrombosis secondary to acute pancreatitis. Hepatobiliary Pancreat Dis Int. 2020;19(2):116‐121. [DOI] [PubMed] [Google Scholar]
  • 6.Butler JR, Eckert GJ, Zyromski NJ, et al. Natural history of pancreatitis-induced splenic vein thrombosis: a systematic review and meta-analysis of its incidence and rate of gastrointestinal bleeding. HPB (Oxford). 2011;13(12):839‐845. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Nawacki Ł, Matykiewicz J, Stochmal E, et al. Splanchnic vein thrombosis in acute pancreatitis and its consequences. Clin Appl Thromb Hemost. 2021;27:10760296211010260. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Nadkarni NA, Khanna S, Vege SS. Splanchnic venous thrombosis and pancreatitis. Pancreas. 2013;42(6):924‐931. [DOI] [PubMed] [Google Scholar]
  • 9.Rebours V, Boudaoud L, Vullierme MP, et al. Extrahepatic portal venous system thrombosis in recurrent acute and chronic alcoholic pancreatitis is caused by local inflammation and not thrombophilia. Am J Gastroenterol. 2012;107(10):1579‐1585. [DOI] [PubMed] [Google Scholar]
  • 10.Harris S, Nadkarni NA, Naina HV, et al. Splanchnic vein thrombosis in acute pancreatitis: a single-center experience. Pancreas. 2013;42(8):1251‐1254. [DOI] [PubMed] [Google Scholar]
  • 11.Głuszek S, Nawacki Ł, Matykiewicz J, et al. Severe vascular complications of acute pancreatitis. Pol Przegl Chir. 2015;87(10):485‐490. [DOI] [PubMed] [Google Scholar]
  • 12.Practice guidance for diagnosis and treatment of pancreatitis-related splanchnic vein thrombosis (Shenyang, 2020). J Dig Dis. 2021;22(1):2‐8. [DOI] [PubMed] [Google Scholar]
  • 13.Duran M, Pohl E, Grabitz K, et al. The importance of open emergency surgery in the treatment of acute mesenteric ischemia. World J Emerg Surg. 2015;10:45.. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Anis FS, Adiamah A, Lobo DN, et al. Incidence and treatment of splanchnic vein thrombosis in patients with acute pancreatitis: a systematic review and meta-analysis. J Gastroenterol Hepatol. 2022;37(3):446-454. [DOI] [PubMed] [Google Scholar]
  • 15.Hajibandeh S, Hajibandeh S, Agrawal S, et al. Anticoagulation versus no anticoagulation for splanchnic venous thrombosis secondary to acute pancreatitis: do we really need to treat the incidental findings? Pancreas. 2020;49(9):e84‐ee5. [DOI] [PubMed] [Google Scholar]
  • 16.Sissingh NJ, Groen JV, Koole D, et al. Therapeutic anticoagulation for splanchnic vein thrombosis in acute pancreatitis: a systematic review and meta-analysis. Pancreatology. 2022;22(2):235‐243. [DOI] [PubMed] [Google Scholar]
  • 17.Chandan S, Buddam A, Khan SR, et al. Use of therapeutic anticoagulation in splanchnic vein thrombosis associated with acute pancreatitis: a systematic review and meta-analysis. Ann Gastroenterol. 2021;34(6):862‐871. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Zhou J, Zhang H, Mao W, et al. Efficacy and safety of early systemic anticoagulation for preventing splanchnic thrombosis in acute necrotizing pancreatitis. Pancreas. 2020;49(9):1220‐1224. [DOI] [PubMed] [Google Scholar]
  • 19.Garret C, Péron M, Reignier J, et al. Risk factors and outcomes of infected pancreatic necrosis: retrospective cohort of 148 patients admitted to the ICU for acute pancreatitis. United European Gastroenterol J. 2018;6(6):910‐918. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Franco L, Becattini C, Beyer-Westendorf J, et al. Definition of major bleeding: prognostic classification. J Thromb Haemost. 2020;18(11):2852‐2860. [DOI] [PubMed] [Google Scholar]
  • 21.Wang L, Guo X, Xu X, et al. Anticoagulation favors thrombus recanalization and survival in patients with liver cirrhosis and portal vein thrombosis: results of a meta-analysis. Adv Ther. 2021;38(1):495‐520. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Pagliari D, Cianci R, Brizi MG, et al. Anticoagulant therapy in the treatment of splanchnic vein thrombosis associated to acute pancreatitis: a 3-year single-centre experience. Intern Emerg Med. 2020;15(6):1021‐1029. [DOI] [PubMed] [Google Scholar]
  • 23.Clark J, Tobin N, Eaton B, et al. Splanchnic thrombosis in moderate to severe pancreatitis. J Am Coll Surg. 2020;231(4):e139. [Google Scholar]
  • 24.Anderson W, Niccum B, Chitnavis M, et al. Outcomes of anticoagulation for portal and/or splenic vein thrombosis in setting of acute pancreatitis. Am J Gastroenterol. 2017;112:S6. [Google Scholar]
  • 25.Wang Y, Attar BM, Jaiswal P, et al. Thrombosis of splanchnic veins during a first episode of acute pancreatitis: prevalence and outcomes. Gastroenterology. 2017;152(5):S279. [Google Scholar]
  • 26.Toqué L, Hamy A, Hamel JF, et al. Predictive factors of splanchnic vein thrombosis in acute pancreatitis: a 6-year single-center experience. J Dig Dis. 2015;16(12):734‐740. [DOI] [PubMed] [Google Scholar]
  • 27.Hall TC, Garcea G, Metcalfe M, et al. Impact of anticoagulation on outcomes in acute non-cirrhotic and non- malignant portal vein thrombosis: a retrospective observational study. Hepatogastroenterology. 2013;60(122):311‐317. [PubMed] [Google Scholar]
  • 28.Gonzelez HJ, Sahay SJ, Samadi B, et al. Splanchnic vein thrombosis in severe acute pancreatitis: a 2-year, single-institution experience. HPB (Oxford). 2011;13(12):860‐864. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Pribramska V, Vege SS, Trna J, et al. Natural history of splanchnic venous thrombosis in acute pancreatitis: a population-based study. Gastroenterology. 2009;136(5):A541. [Google Scholar]
  • 30.Thejasvin K, Chan SJ, Varghese C, et al. A selective anticoagulation policy for splanchnic vein thrombosis in acute pancreatitis is associated with favourable outcomes: experience from a UK tertiary referral centre. HPB (Oxford). 2022;24(11):1937-1943. [DOI] [PubMed] [Google Scholar]
  • 31.Yang AL, Vege SS. Splanchnic vein thrombosis in acute pancreatitis: a study in 967 recent, direct, consecutive admissions at a tertiary-care institution. Gastroenterology. 2015;148(4):S684. [Google Scholar]
  • 32.Harris S, Naina HVK, Vege SS. Splanchnic vein thrombosis in acute pancreatitis: a single center experience. Pancreas. 2008;37(4):473. [DOI] [PubMed] [Google Scholar]
  • 33.Junare PR, Udgirkar S, Nair S, et al. Splanchnic venous thrombosis in acute pancreatitis: does anticoagulation affect outcome? Gastroenterology Res. 2020;13(1):25‐31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Muddana V, Easler JJ, Slivka A, et al. Prevalence, risk factors and management of peripancreatic venous thrombosis (PPVT) in acute pancreatitis (AP). Gastroenterology. 2012;142(5):S320. [Google Scholar]
  • 35.Ahmed SU, Rana SS, Ahluwalia J, et al. Role of thrombophilia in splanchnic venous thrombosis in acute pancreatitis. Ann Gastroenterol. 2018;31(3):371‐378. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Hammer HF. An update on pancreatic pathophysiology (do we have to rewrite pancreatic pathophysiology?). Wien Med Wochenschr. 2014;164(3-4):57‐62. [DOI] [PubMed] [Google Scholar]
  • 37.Lasson A, Ohlsson K. Disseminated intravascular coagulation and antiprotease activity in acute human pancreatitis. Scand J Gastroenterol Suppl. 1986;126:35‐39. [DOI] [PubMed] [Google Scholar]
  • 38.Mendelson RM, Anderson J, Marshall M, et al. Vascular complications of pancreatitis. ANZ J Surg. 2005;75(12):1073‐1079. [DOI] [PubMed] [Google Scholar]
  • 39.Golemi I, Salazar Adum JP, Tafur A, et al. Venous thromboembolism prophylaxis using the Caprini score. Dis Mon. 2019;65(8):249‐298. [DOI] [PubMed] [Google Scholar]
  • 40.Gigante A, Barbano B, Sardo L, et al. Hypercoagulability and nephrotic syndrome. Curr Vasc Pharmacol. 2014;12(3):512‐517. [DOI] [PubMed] [Google Scholar]
  • 41.Aksu K, Donmez A, Keser G. Inflammation-induced thrombosis: mechanisms, disease associations and management. Curr Pharm Des. 2012;18(11):1478‐1493. [DOI] [PubMed] [Google Scholar]
  • 42.Levi M, van der Poll T. Inflammation and coagulation. Crit Care Med. 2010;38(2 Suppl):S26‐S34. [DOI] [PubMed] [Google Scholar]
  • 43.Kröner PT, Wallace MB, Raimondo M, et al. Systemic anticoagulation is associated with decreased mortality and morbidity in acute pancreatitis. Pancreatology. 2021;21(8):1428‐1433. [DOI] [PubMed] [Google Scholar]
  • 44.Clair DG, Beach JM. Mesenteric ischemia. N Engl J Med. 2016;374(10):959‐968. [DOI] [PubMed] [Google Scholar]
  • 45.Luo X, Nie L, Wang Z, et al. Transjugular endovascular recanalization of splenic vein in patients with regional portal hypertension complicated by gastrointestinal bleeding. Cardiovasc Intervent Radiol. 2014;37(1):108‐113. [DOI] [PubMed] [Google Scholar]
  • 46.Zheng K, Guo X, Feng J, et al. Gastrointestinal bleeding due to pancreatic disease-related portal hypertension. Gastroenterol Res Pract. 2020;2020:3825186. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.DiMaio CJ. Management of complications of acute pancreatitis. Curr Opin Gastroenterol. 2018;34(5):336‐342. [DOI] [PubMed] [Google Scholar]
  • 48.Scherer RW, Saldanha IJ. How should systematic reviewers handle conference abstracts? A view from the trenches. Syst Rev. 2019;8(1):264. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

sj-docx-1-cat-10.1177_10760296231188718 - Supplemental material for Anticoagulation Therapy for Splanchnic Vein Thrombosis Associated With Acute Pancreatitis: A Systematic Review and Meta-Analysis
Click here for additional data file. (309.1KB, docx)

Supplemental material, sj-docx-1-cat-10.1177_10760296231188718 for Anticoagulation Therapy for Splanchnic Vein Thrombosis Associated With Acute Pancreatitis: A Systematic Review and Meta-Analysis by Yuhang Yin, Le Wang, Fangbo Gao, Lei Liu and Xingshun Qi in Clinical and Applied Thrombosis/Hemostasis

Articles from Clinical and Applied Thrombosis/Hemostasis are provided here courtesy of SAGE Publications

RESOURCES