Role of thrombophilia in splanchnic venous thrombosis in acute pancreatitis

Introduction

Acute pancreatitis (AP) is an acute inflammatory process of the pancreas with wide clinical presentations, ranging from mild discomfort to catastrophic debility and even death [1,2]. The inflammatory process may be limited to the pancreas or may involve regional or distant organ systems, leading to local or systemic complications [1,2]. The arterial and venous complications of AP, including thrombosis and pseudoaneurysms, are an important cause of morbidity and mortality [3,4].

Splanchnic venous system thrombosis (SVT) is one of the common vascular complications of AP [5]. There is wide variation in the incidence of SVT in AP between different studies, with reported frequencies ranging from as low as 1.8% to as high as 36.5% [6-9]. Several studies have looked at the pathophysiological mechanisms associated with SVT [8-10]. Factors that have been implicated are pancreatic and peri-pancreatic edema, inflammatory cellular infiltration, direct injury to blood vessels and fluid collection compressing the vessels and causing stasis of blood flow. Peri-pancreatic inflammatory changes or fluid collecstions create a milieu that promotes the development of SVT [9,11].

Thrombophilia has also been implicated as a cause of SVT in AP but this has not been proven conclusively. Deficiencies of natural anticoagulants, such as proteins C and S, have been found in patients with AP [12] and can lead to SVT [13]. There are insufficient data in the literature regarding the role of other coagulation parameters in AP, such as anticardiolipin antibody, lupus anticoagulant, and anti-β2 glycoprotein I (β2GPI) [14].

Moreover, there is a paucity of literature on the natural history of SVT in AP, with the majority of studies reporting that complications directly related to SVT are infrequent and that most of the patients remain asymptomatic with regard to splanchnic vein thrombosis [8,9]. In addition, spontaneous recanalization of a thrombosed splanchnic venous system has been reported in up to 30% of patients [15].

We conducted this prospective study to evaluate the frequency, risk factors and extent of SVT in patients with AP and its impact on outcome parameters including the presence or absence of necrosis, organ failure, need for intervention, and mortality. These patients also underwent a detailed analysis of coagulation parameters to evaluate the role of thrombophilia in the causation of SVT in AP.

Patients and methods

Results

One hundred eleven patients were prospectively selected, of whom 15 were excluded (recurrent pancreatitis in 9, coexistent rheumatoid arthritis in 2, coexistent chronic liver diseases in 2, coexistent ulcerative colitis in 1, and 1 patient where CT could not be done). The remaining 96 patients (76% male; mean age 31.85±13.34 years, range 15-77 years) were evaluated with CECT of the abdomen for the presence of SVT. The most common etiology of AP was alcohol (45.8%), followed by gallstone disease (32.3%), whereas 13.5% of patients had idiopathic AP. All the patients had abdominal pain on presentation.

SVT

SVT was present in 26 (27.1%) patients. A single vessel, two vessels and all three vessels were involved in 19 (73.1%), 3 (11.5%), and 4 (15.4%) patients, respectively. The most commonly involved vessel was the splenic vein in 22 (84.6%) patients, followed by the portal vein in 11 (42.3%) patients and the superior mesenteric vein in 4 (15.3%). Table 1 shows the pattern of vascular thrombosis.

Table 1 Patterns of splanchnic vascular thrombosis (SVT)

Comparison of patients with and without SVT

When patients with SVT were compared to those without, there was no significant difference between the two groups with regard to age, sex, duration of hospitalization or etiology of AP. However, on clinical presentation, a lump or pleural effusion were present in significantly more patients in the SVT group compared with the no-SVT group (Table 2). Laboratory parameters were comparable between the two groups. When the outcome variables were compared between the two groups, the presence of necrotizing pancreatitis, CTSI >6, and MCTSI >6 were significantly associated with SVT. There was no significant association of SVT with BISAP >2, presence of SIRS, presence and number of organ failure, severity, need for intervention, or mortality (Table 3).

Table 2 Comparison of clinical features in patients of acute pancreatitis with or without splanchnic vascular thrombosis (SVT)

Table 3 Comparison of outcome variable between patients with and without splanchnic vascular thrombosis (SVT)

Coagulation workup

A detailed procoagulation workup, including protein C, protein S, AT III, lupus anticoagulant, anticardiolipin antibody, β2GPI, PT, aPTT, fibrinogen, and d-dimer assay, was performed in 42 patients, 18 with SVT and 24 without. Samples for coagulation studies were collected within 4 weeks of the onset of pain in 21 patients (early group) and after 4 weeks in 21 patients (late group). FVL mutation analysis was performed in 33 patients, 19 without SVT and 14 with SVT. During the procoagulant workup, at least one abnormality was found in 25 (59.5%) patients. Protein C deficiency was found in 10 patients (23.8%), protein S deficiency in 14 patients (33.3%), and AT III deficiency in 13 patients (31%). β2GPI was positive in 2 patients (4.8%) and lupus anticoagulant was positive in 5 patients (11.9%). No patient had an anticardiolipin antibody abnormality. Of the 33 patients undergoing FVL mutation analysis, 2 (6.1%) were found to have a heterozygous mutation. Among the global coagulation parameters, some abnormality was found in most of the patients. Prolonged PT was found in 41 of the 42 patients (97.6%) while prolonged aPTT was found in only 2 patients (4.8%). Fibrinogen level was elevated in 23 patients (54.8%), while d-dimer testing was positive in 36 patients (85.7%). More patients in the early group (coagulation study within 4 weeks of pain onset) had abnormalities in AT III, fibrinogen, and d-dimer compared to the late group (coagulation study >4 weeks after pain onset) (Table 4).

Table 4 Coagulation parameters; early (<4 weeks) vs. late (>4 weeks)

When patients with normal procoagulants were compared to those with an abnormal procoagulant workup, there was no significant difference between the groups with regard to age, sex, duration of hospitalization, etiology or any of the outcome variables (Table 5).

Table 5 Comparison of outcome variables between patients with normal and abnormal procoagulant results

Coagulation parameters and SVT

Among the 26 patients with SVT, 18 underwent a procoagulant workup, as did 24 patients without SVT. Among the patients with SVT an abnormal result was found in 8 (44.4%) patients, while of the 24 patients without SVT 7 (29.2%) had abnormal results. The difference between the two groups was not statistically significant (Table 6).

Table 6 Coagulation parameters in patients with acute pancreatitis with or without splanchnic vascular thrombosis (SVT)

Discussion

Thrombosis involving the splanchnic veins is a known and common complication of AP. The reported incidence of SVT in previous studies ranged from 1.8-36.5% [6-9,17-19]. In our cohort of 96 patients, SVT was found in 26 (27.1%) patients. The most commonly involved vessel was the splenic vein in 22 of 26 (84.6%), followed by the portal vein in 11 (42.3%). The superior mesenteric vein was the least common vessel to be involved, with thrombosis observed in only 4 (15.3%) patients. This pattern of vascular involvement is consistent with previous studies [8,9]. Almost all the previous studies reported the splenic vein to be the most commonly involved vessel.

We did not find any significant correlation between the presence or absence of SVT and patients’ age, sex, or the etiology of AP. A study by Easler et al also found no significant correlation between these two groups and age, sex or etiological factors [9]. In contrast, Zhou et al found that male sex was associated with a higher probability of SVT and that an idiopathic cause was more common in patients without SVT. However, on multiple logistic regression analysis these factors did not differ significantly between the two groups [20]. We also found that presence of a palpable abdominal lump was significantly associated with the presence of SVT (50% vs. 28.6%; P=0.049). A palpable lump in case of AP is usually due to an inflammatory mass or a collection. The presence of pancreatic or peri-pancreatic collection and inflammation has been postulated as a cause of SVT [4,11]. Pleural effusion was also seen in a significantly greater number of patients with SVT, which may be related to a higher degree of local inflammation in AP patients with pleural effusion.

Pancreatic and peri-pancreatic necrosis and the associated inflammatory process cause vascular inflammation and infiltration of vessels by inflammatory cells, inducing stasis leading to thrombus formation [8,9]. In our study, consistently with this hypothesis, the presence of necrotizing pancreatitis was significantly associated with a greater probability of developing SVT. The percentage of necrosis was higher in patients with SVT compared to those without. Previous studies have reported that a higher CTSI is associated with a greater probability of SVT [18,20]. In our study, severe pancreatitis, defined by a CTSI >6, was seen in a significantly greater number of patients with SVT compared to those without SVT (76.9% vs. 47.1%, P=0.009).

We found no significant association between the need for any intervention, such as surgery or endotherapy, in patients with SVT compared to those without. Though the need for percutaneous catheter drainage was more common in patients with SVT compared to those without SVT, the difference did not reach the statistical significance. Earlier studies have reported that patients with SVT have longer hospital stays compared to those without [6-9,17-19]. Mortality as a direct consequence of SVT has not been reported in patients with AP, although SVT has a high mortality outside the setting of AP [21,22]. We also found no significant difference in mortality between the two groups.

Ours is a large study as far as the coagulation workup in AP is concerned. Among the procoagulant workup, at least one abnormality was present in more than half of the patients. Previous studies have reported that systemic activation of the coagulation system commonly occurs in over half of the critically ill patients with severe AP [12]. Proteins C and S, and AT III deficiencies were found in 23.8%, 33.3%, and 31.0% of patients respectively, similar to previous studies [12,23-25]. Deficiency of these natural anticoagulants occurs as part of the systemic inflammatory response. Systemic inflammation causes downregulation of hepatic synthetic functions, leading to a deficiency of these proteins [12]. The high number of patients with such a deficiency in our study can be explained by the large number of patients who had SIRS. Ours is the largest study looking into lupus anticoagulant, β2GPI, anticardiolipin antibody, and FVL mutation analysis. These procoagulant parameters have previously been evaluated in only a small number of AP patients with SVT [7].

We found no significant correlation between patients having normal or abnormal procoagulant and any of the outcome variables. This is in contrast to previous studies, where higher mortality and multiple organ failure were found to be associated with abnormal procoagulation parameters [12,26]. We found prolongation of PT in most (97.6%) of the patients and prolonged aPTT in only few (4.8%) patients, consistent with previous studies [27,28]. Prolongation of PT may be secondary to disseminated intravascular coagulation (DIC), associated with severity inflammatory processes such as pancreatitis; or it may be secondary to impaired hepatic synthetic function caused by cytokine as a part of SIRS. In DIC, the level of fibrinogen is decreased and that of d-dimer is elevated. This is in contrast to our study, as we found increased fibrinogen levels in 54.8% of patients. However, an increased fibrinogen level can be due to the fact that fibrinogen is an acute phase reactant, whose level may be increased in acute inflammatory conditions like AP. Consistently with this, in our study we found that high fibrinogen levels were found in a significantly higher number of patients whose coagulation study was performed early in the course of AP (<4 weeks) compared to those whose study was performed later (>4 weeks). The increased level of d-dimer (in 87.5% of patients) indicates fibrinolysis, suggesting that fibrinogen production and fibrinolysis go hand in hand in AP. This is consistent with our finding that a significantly higher number of patients had increased d-dimer levels when their coagulation study was done early rather than late in the course of AP.

Compared to the patients without SVT, patients with SVT did not show any significant difference in the results of their coagulation studies. In fact, an overall abnormality of procoagulant parameters was seen more commonly in the group without SVT. This was probably due to a systemic inflammatory response and cytokine production in AP, causing impaired hepatic synthetic function (protein C and S deficiency) or consumption coagulopathy secondary to diffuse microvascular thrombi. One of our patients with positive lupus anticoagulant had involvement of all three vessels in thrombosis. Even if we assume that the presence of positive lupus anticoagulant caused this extensive thrombosis, two other patients in whom all three vessels were involved had no abnormality in procoagulant parameters, indicating that extent of vascular thrombosis does not have any correlation with normal or abnormal coagulation parameters.

Follow-up ultrasound Doppler imaging showed that SVT resolved in most of the patients within 1 year without the use of any anticoagulation therapy. Studies by Gonzalez et al and Harris et al reported that one third of patients with SVT had resolution of thrombus without anticoagulants [7,8]. However, some authors have reported that anticoagulant treatment is useful and should be applied in patients with SVT [29]. Complications in AP, particularly those related to SVT, are infrequent. We found no complication specifically related to SVT, such as variceal bleeding or intestinal ischemia.

Our study had certain important limitations. There was a higher proportion of patients with acute necrotizing pancreatitis in our study; this is because our center is a tertiary care referral institution in North India where a majority of sick cases are referred, whereas patients with mild pancreatitis are usually treated at peripheral centers. In addition, the study population was not homogeneous—being mainly male (76%) and with alcohol-induced pancreatitis (45.8%), just a minority of cases having idiopathic pancreatitis—which could have led to bias [20]. Although none of the patients with alcohol-induced pancreatitis had overt liver disease, subclinical chronic liver disease could have influenced their real coagulation status. Other limitations are that few patients with SVT were followed up (7/26 patients) and for shorter durations, while the follow up used Doppler imaging, which always remains an operator-dependent method. Finally, there are no established guidelines supporting the use of Doppler imaging at regular intervals for assessing the status of SVT and our selection of a cutoff of 4 weeks after the onset of pain for assessing coagulation parameters was arbitrary.

In conclusion, SVT is a common finding in AP and is significantly associated with the presence of necrotizing pancreatitis. It does not significantly impact the outcome of the disease, it is not related to normal or abnormal blood coagulation parameters, and in most patients it resolves without any intervention.