Chapter 26: Splenic Artery Embolization for Management of Hypersplenism and Portal Hypertension Partial splenic embolization (PSE) is a technique that provides many benefits in the setting of hypersplenism and portal hypertension (PHT); these include decreasing the incidence of variceal bleeding and hepatic encephalopathy while increasing liver protein production, platelet counts, and white blood cell (WBC) counts. A recent review article by Koconis et al nicely documents the history, efficacy, and safety of PSE.1 Splenic embolization using autologous clot was first described in 1973 by Maddison for treatment of hemorrhage from gastrointestinal varices.2 Early in the experience with splenic embolization, there were unacceptably high rates of splenic abscess, pneumonia, sepsis, splenic rupture, and death. This was predominately because of very aggressive embolization techniques and the lack of antibiotic prophylaxis. These severe complications limited the popularity of the procedure. Improved clinical results were obtained with the addition of decreased embolization volume, antibiotic prophylaxis, and adequate pain control as initially described by Spigos et al.3 The clinical utility of PSE was slow to develop because of the concurrent development and popularity of both transjugular intrahepatic portosystemic shunts (TIPS) and the subsequent development of balloon-occluded retrograde transvenous obliteration (BRTO) in the management of PHT and varices. However, in the correct clinical context, PSE is a safe and effective procedure that should be considered for the management of patients with hypersplenism and PHT. Chikamori et al evaluated 37 patients who underwent PSE for hypersplenism.4 The authors found that the wedged hepatic venous pressure before PSE averaged 39 ± 10 cm H2O, which was reduced to 33 ± 8 cm H2O after PSE. Additionally, the flow volumes in the splenic vein decreased from 477 ± 200 mL/min before PSE to 319 ± 187 mL/min after PSE. However, in this study, there was no significant change in the flow volume within the portal vein (PV).4 This last finding is likely due to the increase in mesenteric arterial flow after PSE, which results in increased mesenteric venous flow, offsetting the decreased splenic venous flow and thus maintaining portal venous flows.5 The reported ideal degree of splenic devascularization from PSE is widely variable throughout the literature, ranging from 20% to 70%. Sangro et al recommend a goal of 60% to 70% devascularization to maintain a good response without adverse outcomes. In their experience, there was a relapse of hypersplenism in patients with less than 50% devascularization.6 Other investigators have had success with 30% to 40% embolization, with the plan to repeat PSE as needed to control the underlying sequelae.7 Similar to patients undergoing splenectomy, it may be worthwhile to administer Haemophilus influenzae, pneumococcal, and meningococcal vaccination before PSE in case there is an inadvertent high percentage of splenic embolization. In a review of patients with splenic trauma, Nakae et al compared 24 patients who underwent splenectomy with 34 patients who underwent splenic preservation treatment (embolization, partial splenectomy, or splenorrhaphy). They found that there was no significant difference in the levels of IgM or specific IgG antibodies against 14 types of Streptococcus pneumoniae capsular polysaccharide between the two groups. This suggests that embolization is as detrimental to immune function as splenectomy and that vaccination should be performed.8 In terms of embolic agent choice, Gelfoam (Upjohn, Kalamazoo, Michigan) is the embolic agent most frequently described in the literature.1 Coils, medium to large polyvinyl alcohol (PVA) particles, N-butyl cyanoacrylate (NBCA), or absolute alcohol can also be used depending on operator preference and goals of the embolization. One advantage of NBCA and absolute alcohol as embolic agents is that they do not rely on the patient’s coagulation status. There is considerable variability in the literature with regard to how distal the embolization should be; more distal embolizations (i.e., distal to the splenic hilum) may result in higher complication rates9 but may also prevent development of collaterals that can lead to a recurrence of hypersplenism.10 In a recent meta-analysis, Schnüriger et al found that the main minor complication from more distal embolization was a higher rate of segmental splenic infarction.11 In a separate small series of children with hypersplenism from thalassemia, the lower pole splenic artery was selectively embolized, which was thought to help reduce the development of left-sided pleural effusions.12 Partial splenic embolization is effective in decreasing the rate of variceal bleeding in patients with PHT, regardless of the etiology. Pålsson et al followed a cohort of 26 severely ill patients with bleeding esophageal varices and thrombocytopenia.13 These patients were treated a total of 52 times over the course of 20 plus years with PSE, using Gelfoam as the embolic agent with a splenic devascularization goal of 30% to 40%. The average number of esophageal variceal bleeding episodes per patient decreased from 4.3 ± 2.9 before PSE to 1.1 ± 1.7 after PSE (P< 0.001).13 In the aforementioned review article by Koconis et al, there were four studies that had a combined 50 patients with an average of 2.4 variceal bleeding episodes per year that decreased to an average of 0.48 bleeding episodes per year, a reduction of 80%.1 In a study by Ohmagari et al, 17 patients with portal hypertensive gastropathy were treated with PSE and compared with 13 control participants. When compared with control participants, PSE resulted in an 11% reduction in gastric mucosal hemoglobin content (P< 0.01). There was also a 71% rate of improvement in portal hypertensive gastropathy in the PSE group compared with 8% in the control group at follow-up endoscopy (P< 0.05).14 Partial splenic embolization has been found to be effective at improving synthetic liver function. In the review article by Koconis et al, there were two studies that had a combined 40 patients who had improved total cholesterol levels, total protein levels, albumin levels, and prothrombin time that persisted for more than 1 year after PSE.1 The etiology for this improvement remains unclear but may be related to increased mesenteric arterial flow, with subsequent improvement in the nutritional levels in the PV. Platelet and WBC counts also improve after PSE in patients with hypersplenism. Alzen et al described their experience with 17 patients ranging in age from 1 to 31 years in which hypersplenism was defined as an enlarged spleen in a patient with pancytopenia.15 The main etiologies for splenomegaly in this patient population were PV thrombosis (five patients) and cystic fibrosis with concomitant cirrhosis (three patients). All patients were embolized with 150- to 355-micron PVA particles with a goal devascularization of 30% to 60%. Ten of the 17 (59%) patients developed a temporary left-sided effusion or ascites. The mean platelet count increased from 51,000/µL to 275,000/µL two weeks after the procedure; this increase was likely due to decreased splenic sequestration. Additionally, the mean WBC count increased from 2800 to 8200/µL two weeks after the procedure, potentially because of decreased leukocyte pooling in the spleen.15 Nagata et al treated 15 patients with hypersplenism, with NBCA diluted 3 to 1 or 4 to 1 with ethiodized oil. Goal devascularization in this study was 60 to 80%. These authors reported that the mean platelet count increased from 45,600 to 174,000/µL two weeks after the procedure and remained elevated at 88,100/µL 2 months after the procedure and 97,600/µL 2 years after the procedure.16 This overall improvement in platelet count likely helps control variceal bleeding. Similarly, in a series of 32 patients, N’Kontchou et al demonstrated that platelet counts increased by an average of 185% at 1 month after PSE, with 31 of 32 (97%) patients achieving platelet counts greater than 80,000/µL (only one patient was above that threshold before the procedure).17 At 6 months, there was still an average increase of 95% compared with baseline counts, with 20 of 32 patients (63%) above 80,000/µL. Additionally, average leukocyte counts increased by an average of 51% at 1 month and remained elevated by 30% at 6 months. Finally, this study demonstrated usefulness in alleviating pain, which can arise from splenomegaly; all four of the patients in this series with painful splenomegaly treated with PSE experienced resolution of their pain.17 In a prospective study by Ohmoto and Yamamoto, 52 patients with cirrhosis, esophageal varices, and thrombocytopenia were assigned to one of two groups: one treated with endoscopic variceal band ligation (EVL) combined with PSE and the other one treated with EVL alone.18 PSE was carried out to a 60% to 80% devascularization. The combination group had lower relative risks of new varices (RR = 0.390; P = 0.024), variceal bleeding (RR = 0.191; P = 0.021), and death (RR = 0.193; P = 0.012). The long-term cumulative variceal bleeding rate in the combination group was nearly 50% less than in the EVL alone group, with follow-up performed out to 7 years (P = 0.029).18 In a prospective study by Tania et al, 33 patients with esophageal varices were treated with a combination of EVL with PSE and compared with 25 patients who underwent EVL alone.19 The combination therapy group had statistically significantly lower rates of recurrent esophageal varices at 6, 12, and 24 months of 21.1%, 37%, and 58.1%, respectively, compared with the EVL alone group of 58.1%, 70.7%, and 80.4%, respectively.19 BRTO is a method of treating gastric varices through a gastro-renal shunt and can be performed via a transjugular or transfemoral route. The transjugular approach has been associated with a high rate of esophageal variceal development after the procedure (51%–56% at 3 years after the procedure).20 Given that PSE decreases portal venous pressure, it was hypothesized that combining PSE with BRTO might reduce the development of esophageal varices. In a recent study by Chikamori et al, 14 patients were treated with PSE in combination with BRTO to treat gastric varices in patients with a gastrorenal shunt.21 This combination treatment group was compared with 19 patients treated only with BRTO. In the combination therapy group, PSE was performed 7 to 14 days before the obliteration procedure. Gastric varices resolved in 100% of patients in both groups, and there was no significant difference in the 3-year survival rate between the two groups. However, the 3-year cumulative occurrence rate of esophageal varices was only 9% in the combination group but was 45% in the group undergoing BRTO alone (P< 0.05).21 This report demonstrates that PSE added to BRTO may be more effective for the long-term prevention of esophageal varices. Partial splenic embolization added to the obliteration of porto-systemic shunts (PSS) may reduce the degree of hepatic encephalopathy. In a study by Yoshida et al, 25 patients with hepatic encephalopathy were divided into two groups; 14 patients underwent obliteration of PSS followed by PSE, and 11 control patients underwent only obliteration of PSS.22 Serum ammonia levels were lower in the combination group at 6 months, 9 months, 1 year, and 2 years after treatment. Grades of encephalopathy were also lower in the PSE group at 3 months, 6 months, 9 months, and 1 year after treatment.22 Partial splenic embolization has been described as a safe and effective technique to decompress the portal system before live donor liver transplantation, resulting in improved outcomes. Severe PHT increases the risk of intraoperative bleeding and graft hyperperfusion, especially in small-for-size grafts that are often seen in the live donor setting. Additionally, well-developed portal venous collaterals lead to difficulty with perihepatic dissection. In a study by Umeda et al, 60 patients with liver failure from viral or alcoholic etiologies with severe PHT were randomized to two groups: one group underwent proximal splenic artery embolization 12 to 18 hours before live donor liver transplantation, and the other group served as a control.23 Splenic embolization was carried out with coils delivered proximal to the origin of the main pancreatic artery (dorsal pancreatic). There was a statistically significant decrease in portal venous flow and a concurrent increase in hepatic arterial flow in the proximal splenic embolization population. This procedure led to statistically significant decreased operating room time, blood loss, need for transfusion, ascites, posttransplant portal venous velocities, and mortality. Additionally, two patients in the control group developed splenic arterial steal syndrome after transplant and required subsequent proximal splenic embolization.23 Partial splenic embolization is a useful technique in the post– liver transplant population in which either the splenic artery or the gastroduodenal artery steals flow from the transplant hepatic artery. The majority of these cases involve increased flow via the splenic artery as a result of hypersplenism, although steal from the gastroduodenal artery may occur in the setting of superior mesenteric artery stenosis.24 Steal syndrome is a rare complication, occurring in fewer than 6% of transplants, but it results in graft ischemia, particularly of the bile ducts.9 It is often a diagnosis of exclusion after rejection, infection, and vascular thrombosis have been ruled out as a cause of elevated liver enzymes, decreased liver function, and cholestasis. The diagnosis requires careful attention to the dynamic celiac artery angiogram and should be suspected when there is persistent flow in the hepatic arterial branches at the same time that there is portal venous flow from the spleen.24 Other diagnostic criteria for splenic artery steal syndrome include a splenic arterial diameter of 4 mm or larger or 150% larger in diameter than the hepatic artery. The diagnosis can be made as early as 3 weeks or as late as 5.5 years after transplant; however, the majority are diagnosed within 3 months.9 Given the higher risk of infection in the transplant population caused by the administration of immunosuppressive medications, PSE should be performed cautiously with the goal of less than 30% devascularization even if this necessitates repeat intervention.24 In a study by Nussler et al, 29 post-liver transplant patients who were diagnosed with splenic arterial steal syndrome underwent coil embolization of the splenic artery.9 There were rather high rates of adverse outcomes, including PV thrombosis (10.3%), sepsis (46.2%), need for splenectomy (27.6%), and death (17.2%). These complications, however, occurred in the first 15 patients in whom the endovascular technique used distal rather than proximal splenic artery coil embolization. In the latter 14 patients, coil embolization was carried out in the more proximal splenic artery, thus maintaining distal collateral splenic flow, and resulted in no complications and rapid normalization of liver function.9 These results demonstrate the importance of performing the coil embolization in a proximal location for this particular patient population. As with other embolization procedures, patients often develop postembolization syndrome, which consists of fever, nausea, decreased appetite, and abdominal pain that begins 1 to 2 days after the procedure and lasts for up to 2 weeks. Fevers most likely arise from the release of pyrogens by inflammatory cells within the infarcted region and can be reduced by administration of steroids. Abdominal pain can typically be controlled with nonsteroidal antiinflammatory drugs.25 A wide range of minor complications may occur after PSE, including left pleural effusion, ascites, leukocytosis, ileus, and prolonged pain. Depending on the size and symptoms, the pleural effusion and ascites may require drainage. Major complications from PSE more often occur in the setting of a higher percentage of devascularization and include splenic abscess, PV thrombosis, pancreatitis, severe pneumonia, sepsis, bacterial peritonitis, and death. Splenic abscesses often respond to antibiotic management; however, some patients require percutaneous or surgical drainage.25 Finally, the combination of decreased portal flow and increased platelet number after PSE is thought to contribute to the risk of developing PV thrombosis, which is initially managed with anticoagulation.26 In the reported literature, 15 patients with serious complications have been reported; in only 11 of these patients was the degree of embolization reported.1 In eight of these 11 cases, there was 70% or greater devascularization of the spleen. Taking the overall number of patients treated with PSE from published series of at least 10 patients, Koconis et al were able to identify four deaths out of 401 patients, resulting in a 1% mortality rate for the procedure.1 This compares with a 2.8% mortality rate reported for TIPS patients.27 A typical step-by-step procedure, as performed at the authors’ institution, is as follows: 1. Administer antibiotics and consider vaccinations before the procedure. 2. Access the common femoral artery using a micropuncture technique. 3. Exchange the micropuncture system for a 5-Fr vascular sheath. 4. Select the celiac axis with a 5-Fr C2 catheter and perform a celiac artery angiogram. 5. Advance either the C2 catheter or a microcatheter and microwire through the C2 catheter to selectively catheterize the splenic artery and perform a splenic artery angiogram ( 6. Deliver embolic agent of choice to achieve 30% to 60% embolization. The authors’ institution typically uses larger (700–900 micron) particles. 7. Perform a postembolization splenic angiogram ( 8. Remove catheters and perform a common femoral angiogram through the sheath to evaluate for closure device. 9. Remove the femoral sheath and obtain hemostasis with a closure device or manual pressure. 10. Admit the patient overnight for monitoring and pain control. 11. Continue antibiotics for 7 to 14 days after the procedure and a steroid dose pack for 7 days after the procedure; monitor for signs of infection, and maintain adequate pain control. Partial splenic embolization is a safe and effective method for treating patients with hypersplenism and PHT and shows particular promise when used in combination with other therapies. PSE has been shown to be effective in the transplant population as well. See
Introduction
Results and Outcomes
Hemodynamic Results
Clinical Outcome
Combination Therapy
Transplant Population
Complications
Procedural Aspects
Fig. 26.1).
Fig. 26.2).
Conclusion
Table 26.1 for a summary of indications and contraindications.
Thoracic Key
Fastest Thoracic Insight Engine
