CLINICAL STUDY
Comparison of Percutaneous Transhepatic Variceal Embolization (PTVE) Followed by Partial Splenic Embolization versus PTVE Alone for the Treatment of Acute Esophagogastric Variceal Massive Hemorrhage XuHua Duan, MD, Kai Zhang, MM, XinWei Han, MD, JianZhuang Ren, MD, Miao Xu, MM, GuoHao Huang, MM, and MengFan Zhang, MM
ABSTRACT Purpose: To compare the efficacy of percutaneous transhepatic variceal embolization (PTVE) followed by partial splenic embolization (PSE) with that of PTVE alone for the treatment of acute massive hemorrhage of esophagogastric varices in patients with cirrhosis unable to undergo alternative procedures. Materials and Methods: Sixty-five patients with acute variceal massive hemorrhage were retrospectively studied, including 31 who underwent PTVE/PSE and 34 who underwent PTVE and refused PSE. Recurrent bleeding rate, survival rate, postoperative complications, number of days of hospitalization after PTVE, and outcome were evaluated. Peripheral blood cell counts and hemoglobin levels before and at 1 week and 6, 12, and 24 months after intervention were analyzed. Results: Cumulative recurrent bleeding rates at 6, 12, and 24 months after intervention in the PTVE/PSE group were 3.2%, 6.7%, and 13.3%, compared with 20.6%, 36.7%, and 53.6%, respectively, in the PTVE group; the difference at each time point was statistically significant (all P o .01). There were more cases of ascites and portal hypertensive gastropathy after PTVE than after PTVE/PSE (P o .05). Survival rates at 6, 12, and 24 months in the PTVE/PSE group were 100%, 96.8%, and 96.8%, compared with 94.1%, 88.2%, and 82.4%, respectively, in the PTVE group. There were significant differences in peripheral blood cell counts and hemoglobin levels between the PTVE/PSE and PTVE groups at all observed time points (all P o .01). Conclusions: PTVE/PSE not only has long-term efficacy in alleviating hypersplenism, but decreases recurrent bleeding and maintains hepatic reserve in patients with cirrhosis and esophagogastric variceal massive hemorrhage unable to undergo other procedures.
ABBREVIATIONS EVL = endoscopic variceal ligation, TIPS = transjugular intrahepatic portosystemic shunt, HE = hepatic encephalopathy, PTVE = percutaneous transhepatic variceal embolization, PSE = partial splenic embolization, RBC = red blood cell, WBC = white blood cell
Esophagogastric variceal hemorrhage is a common and devastating complication of cirrhosis that results from
From the Department of Interventional Radiology, The First Affiliated Hospital, Zhengzhou University, No. 1 East Jian She Rd., Zhengzhou 450052, Henan Province, People’s Republic of China. Received March 1, 2014; final revision received August 15, 2014; accepted August 20, 2014. Address correspondence to X.H.; E-mail: [email protected] None of the authors have identified a conflict of interest. & SIR, 2014 J Vasc Interv Radiol 2014; 25:1858–1865 http://dx.doi.org/10.1016/j.jvir.2014.08.019
portal hypertension and is associated with significant mortality and morbidity (1,2). As recommended by the American Association for the Study of Liver Diseases guidelines (3,4), a combination of pharmacologic and endoscopic therapies is the most rational approach to the treatment of acute variceal hemorrhage. However, endoscopic therapy is not possible when a clear endoscopic view cannot be obtained as a result of massive bleeding (5,6). Early and multiple recurrence of esophageal and gastric varices can occur even after endoscopic variceal ligation (EVL) (4,7). Creation of a transjugular intrahepatic portal systemic shunt (TIPS) has emerged as an effective modality for
Volume 25
’
Number 12
’
December
’
2014
the prevention of variceal bleeding, even as a rescue therapy after failure of endoscopic approaches (8), even though the use of endoscopic treatment and TIPS creation is restricted in patients with acute variceal hemorrhage who are in critical condition with active bleeding or hemorrhagic shock. The major problems associated with TIPS are inability to alleviate hypersplenism, an increased risk of hepatic encephalopathy (HE), and impaired hepatic function (2,9). Conventional percutaneous transhepatic variceal embolization (PTVE) was introduced for the treatment of ruptured esophageal varices 30 years ago (10). It was recommended for the treatment of acute variceal massive hemorrhage and had a better short-term hemostatic effect than pharmacologic therapy (11). Partial splenic embolization (PSE) appears to be efficacious in reducing episodes of variceal bleeding, improving hematologic parameters, enhancing hepatic protein synthesis, and reducing the severity of HE (12–14). In the present study, we retrospectively analyzed data from 65 patients with acute massive hemorrhage from esophagogastric varices who underwent PTVE or a combined approach comprising PTVE followed by PSE performed 7–10 days later. Our aim was to compare PTVE/PSE with PTVE alone to assess the clinical efficacy of the two treatments for patients with cirrhosis and acute massive hemorrhage of esophagogastric varices unable to undergo alternative procedures in terms of hepatic functional reserve, recurrent bleeding rate, and survival.
MATERIALS AND METHODS Study Design Data from 95 patients with liver cirrhosis and variceal bleeding who underwent interventional treatment in our hospital from January 2010 to February 2012 were retrospectively analyzed. The 95 patients were enrolled for retrospective follow-up. Sixty-five of these patients who completely met the following eligibility criteria were enrolled in the study: (i) diagnosis of liver cirrhosis and portal hypertension without portal vein thrombosis, liver tumor, or cavernous transformation on clinical examination and ultrasonography (US), computed tomography (CT; Fig 1a), or magnetic resonance imaging (ie, with pharmacologic treatment, amount of acute massive bleeding 41,000 mL); (ii) patient had not undergone pericardial devascularization, splenectomy, TIPS creation, or endoscopic therapy for esophagogastric variceal bleeding; (iii) patient was in critical condition that did not permit pericardial devascularization, TIPS creation, or endoscopic therapy after 24 hours of medical therapy had been ineffective because of, eg, physical weakness or hemorrhagic shock; and (iv) patient had no catheterizable gastrorenal shunts and could not be treated by balloon-occluded retrograde transvenous obliteration.
1859
Thirty patients did not fulfill these criteria; of these, 10 had undergone PTVE at another hospital, nine underwent endoscopic therapy, five had a liver tumor, and six were lost to follow-up. Of the 65 included patients, 34 underwent PTVE alone and refused PSE, and 31 underwent PTVE/PSE. This study was approved by the ethics committee of the hospital, and written informed consent was obtained from all patients.
Treatment After transhepatic puncture of a branch of the portal vein under digital subtraction angiographic guidance (Artis zeego; Siemens, Munich, Germany) with a percutaneous transhepatic cholangiography puncture set (Cook, Bloomington, Indiana), a 5-F sheath and a Cobra catheter (Cook) were introduced into the portal vein with a 0.035-inch guide wire (Cook) to measure the portal pressure. After splenoportography to confirm the location of the index varices as well as the feeding vessels and draining veins, the catheter was advanced into the main feeding vessel (left gastric vein). First, to embolize the vascular trunk, coils (3–10 mm 5–12 cm) or microcoils (2–3 mm 2–3 cm; Cook) were injected into the lower esophagus and gastric fundus vessels and all feeding vessels. Angiography was then repeated to assess the extent of variceal obliteration. Next, absolute ethanol was slowly injected to occlude the vascular bed until angiography confirmed that the blood flow in the varices had ceased completely. Splenoportography was repeated to assess the extent of variceal obliteration. If collateral circulation to esophagogastric varices from the portal vein was detected, the aforementioned procedure was repeated until the blood flow in the varices ceased completely (Fig 1b). Finally, the portal pressure was measured again and the puncture tract was embolized with coils. Postinterventional supportive care included treatment for anemia and hypoproteinemia by blood transfusion, reduction of ascites with human serum albumin, and systemic prophylaxis with antibiotic agents for at least 7 days after the procedure. PSE was performed 7–10 days after PTVE in 31 patients in the PTVE/PSE group. The PTVE group was 34 patients who refused PSE after PTVE because they were worried about the postoperative complications of PSE. Briefly, splenic arterial angiography was performed by using a 5.0-F RH or Yashiro catheter (Terumo, Tokyo, Japan) to demonstrate the distribution of splenic arteries and collateral circulation routes. The tip of the catheter was placed at the splenic arteries in the middle or lower pole of the spleen. Polyvinyl alcohol particles (350–560 μm; Cook) were mixed with contrast media and 1 million units of penicillin G and/or 80 mg gentamicin. Under fluoroscopic guidance, the mixture was carefully injected manually through the catheter into the splenic arteries. PSE was performed progressively by means of repeated injections of polyvinyl alcohol under
1860
’
Variceal/Splenic Embolization for Esophagogastric Hemorrhage
Duan et al
’
JVIR
Figure 1. Images from a 59-year-old woman who presented to the emergency department with hematemesis and melena and was treated with PTVE/PSE. (a) CT angiography before PTVE revealed an enlarged gastric varix (arrow), dilated tortuous gastroesophageal varices, and a huge spleen as a result of hypersplenism (asterisk). (b) After occlusion of the enlarged gastric vein (arrow), which was opacified with contrast medium, splenoportography revealed a fine posterior gastric vein feeding the gastroesophageal varices by the collateral circulation. After embolization of this vein with two microcoils, splenoportography revealed complete occlusion of the varices. (c) After PTVE, with 8 days of supportive care, the patient was in hemodynamically stable condition. PSE was performed with the tip of the catheter placed at the splenic arteries in the middle pole of the spleen. (d) Two weeks after the procedure, contrast-enhanced CT showed the splenic infarction ratio to be controlled to approximately 50% of the original splenic volume.
angiography. The splenic infarction ratio was limited to approximately 50%–70% of the original splenic volume (Fig 1c) (15) by two interventional radiologists with 15 and 25 years of experience, who participated in all cases. The exact infarction rate was calculated from a CT scan (Fig 1d) obtained 2 weeks after PSE on a
3.1 Workstation (GE Medical Systems, Milwaukee, Wisconsin) with the use of volumetric analysis software as reported by Zhu et al (15). Supportive care and antipyretic agents for the treatment of postembolization syndrome were administered until the patients recovered and were discharged from the hospital.
Volume 25
’
Number 12
’
December
’
2014
Follow-up Study Sixty-five patients’ vital signs, complications during the procedures, postoperative complications, and the number of days of hospitalization required after PTVE were recorded. Ascites was monitored by US and hypertensive gastropathy by gastroscopy 5–7 days after PTVE or PTVE/PSE. All patients received US; three patients with HE in the PTVE group and two patients with labile hemodynamic findings did not undergo gastroscopy. The patients were followed up in the outpatient clinic at 1 week and 6, 12, and 24 months. Data were collected until recurrent bleeding, death, loss to follow-up, or completion of the study occurred; time to recurrent bleeding and cause of death were recorded. Recurrent bleeding was defined as the presence of “coffee-ground” vomitus, hematemesis, or melena and blood pressure lower than 90 mm Hg after a 24-h period of stable vital signs and hemoglobin levels after PTVE or PSE. Time to recurrence was defined as the time elapsed from the eradication of the acute hemorrhage to recurrent bleeding. Thrombocytopenia, white blood cell (WBC) count, red blood cell (RBC) count, albumin level, hemoglobin level, and Child–Pugh score were determined before PTVE and at 1 week and 6, 12, and 24 months after PTVE alone or PTVE/PSE.
Statistical Analysis Data are shown as means ⫾ standard deviation. Statistical analysis was performed by using SPSS software (version 13.0; IBM, Armonk, New York). The paired Student t test was used to determine statistically significant differences between two groups. One-way analysis of variance was used to compare WBC, RBC, and platelet counts and hemoglobin levels within groups over time. Significance was set at P o .05.
RESULTS Patient Characteristics The characteristics of the patients in the PTVE/PSE and PTVE-alone groups are shown in Table 1. The presence of ascites was determined by review of imaging before treatment. There were no significant differences in any characteristics between the two groups. The main clinical manifestations of hemorrhage were hematemesis and/ or black stools in all cases, and there were early signs of hemorrhagic shock in 44 cases. The amount of bleeding ranged from 1,000 mL to 1,900 mL, with an average of 1,400 mL. Preoperatively, the total transfusion volume of RBCs and plasma was 1,100–1,700 mL, with an average of 1,550 mL.
Outcomes and Complications of the Procedure The technical success rate of PTVE was 100% (N ¼ 65), with 104 varices successfully embolized in 65 patients.
1861
Table 1 . Characteristics of Study Patients Characteristic
PTVE/PSE (n ¼ 31)
PTVE (n ¼ 34)
Sex Male
21
23
Female Age (y)
10 40.1 ⫾ 10.9
11 39.7 ⫾ 9.9
HBV HCV
25 4
26 5
Budd–Chiari syndrome
2
2
0 7.18 ⫾ 1.72
1 7.09 ⫾ 2.11
4 21
6 23
Etiology (n)
Other Child–Pugh score Child–Pugh class A B
6
5
MELD score Ascites (n)
C
25.9 ⫾ 8.4 7
27.2 ⫾ 7.3 9
Follow-up (mo)
32.3 ⫾ 6.7
31.6 ⫾ 5.8
There were no significant differences between groups in any category. Values presented as means ⫾ standard deviation where appropriate. MELD scores ranges from 6 to 40, with higher scores indicating more severe disease. HBV ¼ hepatitis B virus, HCV ¼ hepatitis C virus, MELD ¼ Model for End-Stage Liver Disease, PSE ¼ partial splenic embolization, PTVE ¼ percutaneous transhepatic variceal embolization.
Active bleeding was controlled in all patients, with a 100% hemostasis rate. No varices ruptured during the procedure in the study. Portal vein pressures were measured before and after PTVE in all patients and found to differ significantly, from 22.9 mm Hg ⫾ 4.3 before the procedure to 29.1 mm Hg ⫾ 2.7 after the procedure (P o .001). No cases of pneumothorax, hemopneumothorax, or abdominal bleeding occurred during the procedure. One patient with poorly controlled hemorrhagic shock had a technical complication. In the process of gastric coronary vein embolization, the last coil tail extended to the portal vein as a result of the patient’s irritability for hemorrhagic shock with labile hemodynamic findings. During follow-up, no portal vein thrombosis developed after 5 days of anticoagulant therapy. The splenic infarction ratio was controlled to 60.5% ⫾ 8.7 of the original splenic volume in the PTVE/PSE group.
Side Effects and Complications after the Procedure Complications after treatment are shown in Table 2. There was no significant difference in the incidences of postoperative complications between the two groups with reference to Society of Interventional Radiology classification of complications. There were more cases of fever and abdominal pain in the PTVE/PSE group than in the PTVE group (P ¼ .035, P ¼ .038) and more cases of ascites and portal hypertensive gastropathy in the
1862
’
Variceal/Splenic Embolization for Esophagogastric Hemorrhage
PTVE group than in the PTVE/PSE group (P ¼ .003, P ¼ .014). There was no significant difference in the number of cases of HE or portal vein thrombosis. One case of portal vein thrombosis was found in each group by US 4 days after the procedure and resolved after 5 days of anticoagulant therapy. Two patients experienced HE in the PTVE group; one recovered and the other died 28 days after the procedure after restoration of liver function and alleviation of HE. No patient had a splenic abscess, abdominal bleeding, pulmonary embolism, or any other serious complication of treatment.
Follow-up Follow-up and prognosis are shown in Table 3. In the PTVE/PSE group, 31 patients were followed up at 1 week and 6 months and 30 patients at 12 and 24 Table 2 . Complications after Treatment
Complication
PTVE/PSE PTVE (n ¼ 31) (n ¼ 34) P Value
Fever Abdominal pain
25 23
11 10
.035 .038
Ascites
3
20
.003
Portal hypertensive gastropathy Hepatic encephalopathy
2 0
13 2
.014 NS
Portal vein thrombosis
1
1
NS
NS ¼ not significant, PSE ¼ partial splenic embolization, PTVE ¼ percutaneous transhepatic variceal embolization.
Duan et al
’
JVIR
months; four patients experienced recurrent bleeding during the follow-up period and one died of progressive liver failure 8 months after the procedure. In the PTVE group, 34 patients were followed up at 1 week, 31 at 6 months, 30 at 12 months, and 28 at 24 months; 15 patients experienced recurrent bleeding, including two who died of uncontrolled recurrent bleeding, three who died of progressive liver failure, and one who died of peritonitis. During follow-up, 17 of the 65 patients had variceal recurrent bleeding. CT angiography was performed to determine whether new collateral circulation to esophagogastric varices had been established. If new gastric coronary veins were established, PTVE was undertaken to control recurrent bleeding. If no gastric coronary veins were found, EVL, TIPS creation, or surgery was undertaken. Of the 17 patients with variceal recurrent bleeding, three in the PTVE/PSE group underwent PTVE again and one underwent EVL; in the PTVE group, 10 patients underwent PTVE, two underwent TIPS creation, and one underwent surgery. Of these 17 patients, none died or experienced recurrent bleeding during follow-up. No significant differences between groups were found in the number of hospitalization days after PTVE or the number of patients with complications or who died during follow-up. The time to recurrence in the PTVE/ PSE group was significantly longer than that in the PTVE group (P o .001). The cumulative recurrent bleeding rates at 6, 12, and 24 months were 3.2% (one
Table 3 . Follow-up and Prognosis Characteristic
PTVE/PSE (n ¼ 31)
PTVE (n ¼ 34)
P Value
Hospitalization days after PTVE (d) Recurrence
12.5 ⫾ 4.4 4
11.2 ⫾ 3.4 15
NS .038
Time to recurrence (mo)
9.5 ⫾ 3.9
5.2 ⫾ 2.5
o .001
3.2
20.6
o .001
6.7
36.7
.001
13.3
53.6
.004
Cumulative recurrent bleeding (%) 6 mo 12 mo 24 mo Child–Pugh score 1 wk
10.48 ⫾ 2.05
11.06 ⫾ 1.99
NS
6 mo 12 mo
9.12 ⫾ 1.02 8.19 ⫾ 1.98
9.87 ⫾ 2.16 9.32 ⫾ 2.03
.041 .036
24 mo
7.21 ⫾ 2.37
8.83 ⫾ 2.42
.012
1
6
NS
Death during follow-up Survival (%) 6 mo
100
94.1
NS
12 mo 24 mo
96.8 96.8
88.2 82.4
NS NS
0 1
2 3
NS NS
Cause of death (n) Recurrent bleeding Progressive liver failure Sepsis
0
0
NS
Peritonitis
0
1
NS
Values presented as means ⫾ standard deviation where appropriate. PSE ¼ partial splenic embolization, PTVE ¼ percutaneous transhepatic variceal embolization, NS ¼ not significant.
Volume 25
’
Number 12
’
December
’
2014
1863
of 31), 6.7% (two of 30), and 13.3% (four of 30) in the PTVE/PSE group and 20.6% (seven of 34), 36.7% (11 of 30), and 53.6% (15 of 28) in the PTVE group, respectively. There were significant differences in the cumulative recurrent bleeding rates between the PTVE/PSE and PTVE groups at 6 months (P o .001), 12 months (P o .001), and 24 months (P ¼ .004; Table 3). Survival rates at 6, 12, and 24 months were 100%, 96.8%, and 96.8% in the PTVE/PSE group and 94.1%, 88.2%, and 82.4% in the PTVE group, respectively. However, there were no statistically significant differences in these survival rates between the two groups. Child–Pugh score worsened from 7.18 ⫾ 1.72 to 10.48 ⫾ 2.05 in the PTVE/PSE group and from 7.48 ⫾ 2.05 to 11.06 ⫾ 1.99 in the PTVE group at 1 week after treatment. The scores in the PTVE/PSE group then improved to 9.12 ⫾ 1.02 at 6 months, 8.19 ⫾ 1.98 at 12 months, and 7.21 ⫾ 2.37 at 24 months, all of which were significantly different from those in the PTVE group at the respective time points (P o .05; Table 3).
Chronologic Changes in Peripheral Blood Cell Counts and Hemoglobin Levels Follow-up data on peripheral blood cell counts and hemoglobin levels at 1 week and 6, 12, and 24 months
after treatment are shown in Tables 4 and 5. Before treatment, there was no significant difference in peripheral blood cell counts or hemoglobin levels between the two groups (P 4 .05). After treatment, in the PTVE group, there were no significant differences in peripheral blood cell counts, platelet count, or hemoglobin level over time. In the PTVE/PSE group, WBC and platelet counts were lowest before treatment, reached a peak at 1 week, decreased during the 2-year follow-up period, and were significantly higher (P o .01) than those in the PTVE group at 1 week and 6, 12, and 24 months after treatment; RBC count and hemoglobin level increased significantly at 1 week, reached a peak at 6 months, decreased at 12 and 24 months after treatment, and were significantly higher (P o .01) than in the PTVE group at these time points.
DISCUSSION Although PTVE, endoscopy, TIPS creation, and pericardial devascularization have been used for the treatment of acute variceal hemorrhage for many years, each modality has drawbacks (4,16–18). Smith-Laing et al (19) showed a reduction or delay in recurrent bleeding after percutaneous transhepatic embolotherapy of
Table 4 . Changes in Platelet and WBC Counts Platelets (109/L)
WBCs (109/L)
Time
PTVE/PSE
PTVE
Pretreatment
40.5 ⫾ 11.4 (n ¼ 31)*
39.3 ⫾ 10.4 (n ¼ 34)
2.3 ⫾ 1.2 (n ¼ 31)*
PTVE/PSE
PTVE 2.2 ⫾ 0.9 (n ¼ 34)
42.1 ⫾ 14.8 (n ¼ 34)†
8.5 ⫾ 1.8 (n ¼ 31)
2.3 ⫾ 0.8 (n ¼ 34)†
6 mo 12 mo
93.5 ⫾ 13.5 (n ¼ 31)* 81.9 ⫾ 6.8 (n ¼ 30)*‡
41.9 ⫾ 17.2 (n ¼ 31)† 40.3 ⫾ 13.5 (n ¼ 30)†
6.2 ⫾ 0.9 (n ¼ 31)* 5.3 ⫾ 1.1 (n ¼ 30)*‡
2.5 ⫾ 1.2 (n ¼ 31)† 2.4 ⫾ 0.7 (n ¼ 30)†
24 mo
67.4 ⫾ 12.3 (n ¼ 30)*‡§
40.6 ⫾ 11.7 (n ¼ 28)†
4.3 ⫾ 1.7 (n ¼ 30)*‡§
2.5 ⫾ 1.3 (n ¼ 28)†
1 wk
112.5 ⫾ 14.8 (n ¼ 31)
PSE ¼ partial splenic embolization, PTVE ¼ percutaneous transhepatic variceal embolization, WBC ¼ white blood cell. *P o .01: platelet and WBC counts at 1 wk vs preprocedure and at 6, 12, and 24 mo. † P o .01: platelet and WBC counts in the PTVE/PSE group vs PTVE group. ‡ P o .01: platelet and WBC counts at 6 mo vs 12 mo and 24 mo. § P o .01: platelet and WBC counts at 12 mo vs 24 mo. Table 5 . Changes in RBC Count and Hemoglobin Level RBC (1012/L)
Hb (g/L)
Time
PTVE/PSE
PTVE
PTVE/PSE
PTVE
Preprocedure
2.5 ⫾ 0.4 (n ¼ 31)
2.4 ⫾ 0.5 (n ¼ 34)
72.3 ⫾ 19.2 (n ¼ 31)
71.2 ⫾ 19.4 (n ¼ 34)
1 wk 6 mo
2.6 ⫾ 0.3 (n ¼ 31)* 3.8 ⫾ 0.5 (n ¼ 31)*
2.5 ⫾ 0.7 (n ¼ 34)† 2.6 ⫾ 0.9 (n ¼ 31)†
81.5 ⫾ 11.7 (n ¼ 31)* 97.2 ⫾ 13.6 (n ¼ 31)*
75.3 ⫾ 13.7 (n ¼ 34)† 82.3 ⫾ 19.4 (n ¼ 31)†
12 mo
3.7 ⫾ 0.6 (n ¼ 30)*
2.4 ⫾ 0.8 (n ¼ 30)†
93.5 ⫾ 15.7 (n ¼ 30)*
75.2 ⫾ 16.7 (n ¼ 30)†
24 mo
3.2 ⫾ 0.7 (n ¼ 30)
83.3 ⫾ 14.7 (n ¼ 30)
76.3 ⫾ 15.5 (n ¼ 28)†
*
†
2.7 ⫾ 1.0 (n ¼ 28)
*
PSE ¼ partial splenic embolization, PTVE ¼ percutaneous transhepatic variceal embolization, RBC ¼ red blood cell. *P o .01: in the PTVE/PSE group, RBC count and Hb level preprocedure were significantly lower than those at different time points after treatment. RBC count and hemoglobin level were increased significantly at 1 wk, reached a peak at 6 mo, and declined at 12 and 24 mo after treatment. † P o .01: in the PTVE/PSE group, RBC count and hemoglobin level were significantly higher than in the PTVE group at different time points after treatment.
1864
’
Variceal/Splenic Embolization for Esophagogastric Hemorrhage
gastroesophageal varices. However, 65% of patients experienced recurrent bleeding at a mean of 4.6 months after successful PTVE in that study (19). Treatment with PTVE was first introduced by Lunderquist and Vang (10), but did not achieve widespread clinical acceptance because of the associated higher recurrent bleeding rate and the development of endoscopic treatment (16). However, the long-term recurrent bleeding rate after EVL remains high (4). Bian et al (4) reported that EVL performed to prevent variceal bleeding was associated with a recurrent bleeding rate of almost 30% and a 9%– 19% early recurrent bleeding rate. Variceal embolization by PTVE and variceal ligation by endoscopy are techniques that occlude portosystemic shunts to control bleeding (4,10,11). Gastrorenal shunts and gastric coronary veins are spontaneous portosystemic shunts formed to reduce portal hypertension; however, portal hypertension inevitably deteriorates after these shunts are occluded. Because cirrhosis cannot be effectively controlled, new gastrorenal shunts and gastric coronary veins are then established as a result of worsening portal hypertension (9). Endoscopic therapy is successful in most patients with controlled bleeding; however, in those who continue to experience bleeding and in whom the endoscopic view is obscured as a result of massive bleeding, TIPS creation is highly effective in achieving hemostasis (8,20). In view of the higher mortality rate associated with pericardial devascularization in the acute setting, TIPS creation is the favored rescue procedure for acute uncontrolled variceal hemorrhage (20). Emergent TIPS creation is recommended for the treatment of acute variceal bleeding in patients with moderate or severe liver failure and severe portal hypertension (2,8). Tesdal et al (9) reported cumulative survival rates of 75%, 69%, and 60% in 53 patients at 12, 24, and 48 months, respectively, after TIPS creation and embolotherapy. High Model for End-Stage Liver Disease scores may be associated with the high mortality rate associated with TIPS creation in the study of Tesdal et al (9). The cumulative recurrent bleeding rates in these 53 patients were 16% and 19% at 2 and 4 years after treatment, respectively. However, development of EVL and the implementation of emergent TIPS creation for acute variceal hemorrhage have been limited in China out of economic concerns and the relatively greater hepatic insufficiency in such patients in this country. With advances in embolization materials such as coils and microcoils, application of microcatheters, and improvement of interventional technologies, the high recurrent bleeding rate was decreased by means of PTVE. The cumulative recurrent bleeding rates at 6, 12, and 24 months after PTVE alone in the present study were 20.6%, 36.7%, and 53.6%, respectively, which are much better than those previously reported (19,21). However, in all patients, portal hypertension increased significantly from 22.88 mm Hg ⫾ 4.3 before PTVE to 29.1 mm Hg ⫾ 2.7 after PTVE. Because the deteriorating portal
Duan et al
’
JVIR
hypertension could not be effectively controlled, the number of patients with ascites or portal hypertensive gastropathy after PTVE was greater than that after PTVE/PSE. Apart from the reduction of splenic blood flow (14) and treatment of pancytopenia (15), PSE may be considered for the reduction of portal pressure in patients with hypersplenism (1) and therefore plays an important role in multiple antihemorrhage measures. Approximately 60%–70% of portal venous blood comes from the splenic vein in patients with portal hypertension. PSE decreases reflux in the splenic vein; hence, blood flow in the main portal vein is decreased, as is its pressure (14). Xu et al (14) reported that EVL with PSE decreased blood flow and pressure in the main portal vein to an extent equivalent to the combined effects of splenectomy and devascularization. Alzen et al (22) reported that portal vein pressure was reduced from 28.7 mm Hg ⫾ 7.4 to 24.3 mm Hg ⫾ 5.9 after PSE, representing 30–60% of the splenic volume. In the present study, portal hypertension was significantly increased after PTVE. However, the PTVE/PSE group had statistically significantly lower cumulative recurrent bleeding rates than the PTVE group at 6, 12, and 24 months and a significantly longer time to recurrence. Although changes in portal pressure after PTVE/ PSE were not measured during follow-up, our findings suggest that performing PSE 7–10 days after PTVE can reduce the increased portal pressure caused by PTVE. As a result of postembolization syndrome following PSE, there were more cases of fever and abdominal pain in the PTVE/PSE group than in the PTVE group. These adverse effects were temporary and can mostly be managed conservatively. However, the adverse effects of PTVE, such as ascites and portal hypertensive gastropathy, need to be controlled. Although PTVE was performed 7–10 days earlier than PTVE/PSE, there was no significant difference between the two groups in terms of the number of hospitalization days required after PTVE. Moreover, PSE is much less invasive than traditional splenectomy and can help avoid the risks of surgery intra- and postoperatively (22). There are many advantages concerning PSE, the most important probably being the preservation of the immune function of the spleen and the treatment of leukocytopenia and thrombocytopenia caused by hypersplenism. The results of the present study are in accordance with the peripheral blood cell and hemoglobin level change trends reported by Zhu et al (15) in alleviating hypersplenism. Hu et al (23) achieved cumulative recurrent bleeding rates of 6.6%, 27.7%, 44.1%, and 65.6% at 1, 3, 5, and 8 years after PTVE combined with PSE. With the same treatment, Liu et al (24) achieved cumulative recurrent bleeding rates of 12.5% (six of 47), 24.5% (11 of 45), and 27.9% (12 of 43) at 12, 24, and 36 months, respectively. As a result of poor hepatic function and adverse reactions to PTVE, the patients in the present study
Volume 25
’
Number 12
’
December
’
2014
would not have tolerated the postembolization syndrome caused by concurrent PSE and PTVE, as described by Hu et al (23) and Liu et al (24). Instead, in the PTVE/PSE group, PTVE was followed by PSE after 7–10 days, when the patients were in hemodynamically stable condition and their liver function had improved with supportive care. Although the percentages of our patients with Child–Pugh class C disease and hemodynamic instability before treatment were more than those reported in the studies of Hu et al (23) and Liu et al (24), cumulative recurrent bleeding rates of 3.2%, 6.7%, and 12.3% were achieved at 6, 12, and 24 months in our PTVE/PSE group. We also found that hepatic functional reserve was maintained during follow-up in patients treated with PTVE/PSE compared with PTVE alone. The present study has several limitations. First, the study was neither blinded nor randomized; therefore, the data could have been affected by various potential biases. However, these may have been partially overcome by the independent follow-up and the similar size of the two groups. The procedures studied here are complicated to perform, and the optimal interval between PSE and PTVE is unclear, which might limit the use of this combined method as the first-choice treatment for acute variceal massive bleeding. In addition, the study was retrospective, with a limited number of cases. A prospective randomized controlled study with a large number of subjects is needed to investigate whether PTVE/PSE is a superior treatment for patients with acute variceal massive bleeding with portal hypertension and cirrhosis. In conclusion, PTVE/PSE is a relatively safe and effective method for the treatment of acute massive bleeding from esophagogastric varices in patients with cirrhosis and portal hypertension who cannot be considered for other interventions. PTVE/PSE not only has longterm efficacy in alleviating hypersplenism, but also treats acute variceal massive bleeding and prevents recurrent bleeding. PTVE/PSE should also help maintain hepatic reserve and may improve the prognosis of cirrhosis in these patients.
REFERENCES 1. Kirby JM, Cho KJ, Midia M. Image-guided intervention in management of complications of portal hypertension: more than TIPS for success. Radiographics 2013; 33:1473–1496. 2. Garcia-Pagan JC, Caca K, Bureau C, et al. Early use of TIPS in patients with cirrhosis and variceal bleeding. N Engl J Med 2010; 362:2370–2379. 3. Garcia-Tsao G, Sanyal AJ, Grace ND, Carey W. Prevention and management of gastroesophageal varices and variceal hemorrhage in cirrhosis. Hepatology 2007; 46:922–938.
1865
4. Bian S, Tian XG, Hu JH, Wang GC, Zhang CQ. Percutaneous transhepatic variceal embolization combined with endoscopic ligation for the prevention of variceal rebleeding. J. Dig Dis 2013; 14:388–395. 5. Okada M, Nakashima Y, Kishi T, et al. Percutaneous transhepatic obliteration for massive variceal rectal bleeding. Emerg Radiol 2012; 19: 355–358. 6. Palejwala AA, Smart HL, Hughes M. Multiple pulmonary glue emboli following gastric variceal obliteration. Endoscopy 2000; 32(suppl):S1–S2. 7. Lo GH, Lai KH, Cheng JS, Huang RL, Wang SJ, Chiang HT. Prevalence of paraesophageal varices and gastric varices in patients achieving variceal obliteration by banding ligation and by injection sclerotherapy. Gastrointest Endosc 1999; 49:428–436. 8. Loffroy R, Estivalet L, Cherblanc V, et al. Transjugular intrahepatic portosystemic shunt for the management of acute variceal hemorrhage. World J Gastroenterol 2013; 19:6131–6143. 9. Tesdal IK, Filser T, Weiss C, Holm E, Dueber C, Jaschke W. Transjugular intrahepatic portosystemic shunts: adjunctive embolotherapy of gastroesophageal collateral vessels in the prevention of variceal rebleeding. Radiology 2005; 236:360–367. 10. Lunderquist A, Vang J. Sclerosing injection of esophageal varices through transhepatic selective catheterization of the gastric coronary vein. A preliminary report. Acta Radiol Diagn (Stockh) 1974; 15:546–550. 11. Pereiras R, Viamonte M, Jr., Russell E, LePage J, White P, Hutson D. New techniques for interruption of gastroesophageal venous blood flow. Radiology 1977? 124:313–323. 12. Koconis KG, Singh H, Soares G. Partial splenic embolization in the treatment of patients with portal hypertension: a review of the English language literature. J Vasc Interv Radiol 2007; 18:463–481. 13. Romano M, Giojelli A, Capuano G, Pomponi D, Salvatore M. Partial splenic embolization in patients with idiopathic portal hypertension. Eur J Radiol 2004; 49:268–273. 14. Xu RY, Liu B, Lin N. Therapeutic effects of endoscopic variceal ligation combined with partial splenic embolization for portal hypertension. World J Gastroenterol 2004; 10:1072–1074. 15. Zhu K, Meng X, Qian J, et al. Partial splenic embolization for hypersplenism in cirrhosis: a long-term outcome in 62 patients. Dig Liver Dis 2009; 41:411–416. 16. Rosch J, Keller FS. Transjugular intrahepatic portosystemic shunt: present status, comparison with endoscopic therapy and shunt surgery, and future perspectives. World J Surg 2001; 25:337–345. 17. Tian X, Wang Q, Zhang C, Liu F, Cui Y, Liu J. Modified percutaneous transhepatic variceal embolization with 2-octylcyanoacrylate for bleeding gastric varices: long-term follow-up outcomes. AJR Am J Roentgenol 2011; 197:502–509. 18. Lee JH, Han HS, Kim HA, Koo MY. Long-term results of fundectomy and periesophagogastric devascularization in patients with gastric fundal variceal bleeding. World J Surg 2009; 33:2144–2149. 19. Smith-Laing G, Scott J, Long RG, Dick R, Sherlock S. Role of percutaneous transhepatic obliteration of varices in the management of hemorrhage from gastroesophageal varices. Gastroenterology 1981; 80: 1031–1036. 20. McAvoy NC, Hayes PC. The use of transjugular intrahepatic portosystemic stent shunt in the management of acute oesophageal variceal haemorrhage. Eur J Gastroenterol Hepatol 2006; 18:1135–1141. 21. Sos TA. Transhepatic portal venous embolization of varices: pros and cons. Radiology 1983; 148:569–570. 22. Alzen G, Basedow J, Luedemann M, Berthold LD, Zimmer KP. Partial splenic embolization as an alternative to splenectomy in hypersplenism– single center experience in 16 years. Klin Padiatr 2010; 222:368–373. 23. Hu YB, Feng GS, Liang HM, et al. Long term effects of interventional devascularization in treating cirrhosis patients with portal hypertension. Chin J Radiol 2009; 43:853–857. 24. Liu W, Chen GS, Sun HL, et al. Comparison of interventional devascularization and surgical devascularization in treatment of upper gastrointestinal hemorrhage for portal hypertensive. Chin J Radiol 2009; 43: 748–752.
Comparison of Percutaneous Transhepatic Variceal Embolization (PTVE) Followed by Partial Splenic Embolization versus PTVE Alone for the Treatment of Acute Esophagogastric Variceal Massive Hemorrhage XuHua Duan, MD, Kai Zhang, MM, XinWei Han, MD, JianZhuang Ren, MD, Miao Xu, MM, GuoHao Huang, MM, and MengFan Zhang, MM
ABSTRACT Purpose: To compare the efficacy of percutaneous transhepatic variceal embolization (PTVE) followed by partial splenic embolization (PSE) with that of PTVE alone for the treatment of acute massive hemorrhage of esophagogastric varices in patients with cirrhosis unable to undergo alternative procedures. Materials and Methods: Sixty-five patients with acute variceal massive hemorrhage were retrospectively studied, including 31 who underwent PTVE/PSE and 34 who underwent PTVE and refused PSE. Recurrent bleeding rate, survival rate, postoperative complications, number of days of hospitalization after PTVE, and outcome were evaluated. Peripheral blood cell counts and hemoglobin levels before and at 1 week and 6, 12, and 24 months after intervention were analyzed. Results: Cumulative recurrent bleeding rates at 6, 12, and 24 months after intervention in the PTVE/PSE group were 3.2%, 6.7%, and 13.3%, compared with 20.6%, 36.7%, and 53.6%, respectively, in the PTVE group; the difference at each time point was statistically significant (all P o .01). There were more cases of ascites and portal hypertensive gastropathy after PTVE than after PTVE/PSE (P o .05). Survival rates at 6, 12, and 24 months in the PTVE/PSE group were 100%, 96.8%, and 96.8%, compared with 94.1%, 88.2%, and 82.4%, respectively, in the PTVE group. There were significant differences in peripheral blood cell counts and hemoglobin levels between the PTVE/PSE and PTVE groups at all observed time points (all P o .01). Conclusions: PTVE/PSE not only has long-term efficacy in alleviating hypersplenism, but decreases recurrent bleeding and maintains hepatic reserve in patients with cirrhosis and esophagogastric variceal massive hemorrhage unable to undergo other procedures.
ABBREVIATIONS EVL = endoscopic variceal ligation, TIPS = transjugular intrahepatic portosystemic shunt, HE = hepatic encephalopathy, PTVE = percutaneous transhepatic variceal embolization, PSE = partial splenic embolization, RBC = red blood cell, WBC = white blood cell
Esophagogastric variceal hemorrhage is a common and devastating complication of cirrhosis that results from
From the Department of Interventional Radiology, The First Affiliated Hospital, Zhengzhou University, No. 1 East Jian She Rd., Zhengzhou 450052, Henan Province, People’s Republic of China. Received March 1, 2014; final revision received August 15, 2014; accepted August 20, 2014. Address correspondence to X.H.; E-mail: [email protected] None of the authors have identified a conflict of interest. & SIR, 2014 J Vasc Interv Radiol 2014; 25:1858–1865 http://dx.doi.org/10.1016/j.jvir.2014.08.019
portal hypertension and is associated with significant mortality and morbidity (1,2). As recommended by the American Association for the Study of Liver Diseases guidelines (3,4), a combination of pharmacologic and endoscopic therapies is the most rational approach to the treatment of acute variceal hemorrhage. However, endoscopic therapy is not possible when a clear endoscopic view cannot be obtained as a result of massive bleeding (5,6). Early and multiple recurrence of esophageal and gastric varices can occur even after endoscopic variceal ligation (EVL) (4,7). Creation of a transjugular intrahepatic portal systemic shunt (TIPS) has emerged as an effective modality for
Volume 25
’
Number 12
’
December
’
2014
the prevention of variceal bleeding, even as a rescue therapy after failure of endoscopic approaches (8), even though the use of endoscopic treatment and TIPS creation is restricted in patients with acute variceal hemorrhage who are in critical condition with active bleeding or hemorrhagic shock. The major problems associated with TIPS are inability to alleviate hypersplenism, an increased risk of hepatic encephalopathy (HE), and impaired hepatic function (2,9). Conventional percutaneous transhepatic variceal embolization (PTVE) was introduced for the treatment of ruptured esophageal varices 30 years ago (10). It was recommended for the treatment of acute variceal massive hemorrhage and had a better short-term hemostatic effect than pharmacologic therapy (11). Partial splenic embolization (PSE) appears to be efficacious in reducing episodes of variceal bleeding, improving hematologic parameters, enhancing hepatic protein synthesis, and reducing the severity of HE (12–14). In the present study, we retrospectively analyzed data from 65 patients with acute massive hemorrhage from esophagogastric varices who underwent PTVE or a combined approach comprising PTVE followed by PSE performed 7–10 days later. Our aim was to compare PTVE/PSE with PTVE alone to assess the clinical efficacy of the two treatments for patients with cirrhosis and acute massive hemorrhage of esophagogastric varices unable to undergo alternative procedures in terms of hepatic functional reserve, recurrent bleeding rate, and survival.
MATERIALS AND METHODS Study Design Data from 95 patients with liver cirrhosis and variceal bleeding who underwent interventional treatment in our hospital from January 2010 to February 2012 were retrospectively analyzed. The 95 patients were enrolled for retrospective follow-up. Sixty-five of these patients who completely met the following eligibility criteria were enrolled in the study: (i) diagnosis of liver cirrhosis and portal hypertension without portal vein thrombosis, liver tumor, or cavernous transformation on clinical examination and ultrasonography (US), computed tomography (CT; Fig 1a), or magnetic resonance imaging (ie, with pharmacologic treatment, amount of acute massive bleeding 41,000 mL); (ii) patient had not undergone pericardial devascularization, splenectomy, TIPS creation, or endoscopic therapy for esophagogastric variceal bleeding; (iii) patient was in critical condition that did not permit pericardial devascularization, TIPS creation, or endoscopic therapy after 24 hours of medical therapy had been ineffective because of, eg, physical weakness or hemorrhagic shock; and (iv) patient had no catheterizable gastrorenal shunts and could not be treated by balloon-occluded retrograde transvenous obliteration.
1859
Thirty patients did not fulfill these criteria; of these, 10 had undergone PTVE at another hospital, nine underwent endoscopic therapy, five had a liver tumor, and six were lost to follow-up. Of the 65 included patients, 34 underwent PTVE alone and refused PSE, and 31 underwent PTVE/PSE. This study was approved by the ethics committee of the hospital, and written informed consent was obtained from all patients.
Treatment After transhepatic puncture of a branch of the portal vein under digital subtraction angiographic guidance (Artis zeego; Siemens, Munich, Germany) with a percutaneous transhepatic cholangiography puncture set (Cook, Bloomington, Indiana), a 5-F sheath and a Cobra catheter (Cook) were introduced into the portal vein with a 0.035-inch guide wire (Cook) to measure the portal pressure. After splenoportography to confirm the location of the index varices as well as the feeding vessels and draining veins, the catheter was advanced into the main feeding vessel (left gastric vein). First, to embolize the vascular trunk, coils (3–10 mm 5–12 cm) or microcoils (2–3 mm 2–3 cm; Cook) were injected into the lower esophagus and gastric fundus vessels and all feeding vessels. Angiography was then repeated to assess the extent of variceal obliteration. Next, absolute ethanol was slowly injected to occlude the vascular bed until angiography confirmed that the blood flow in the varices had ceased completely. Splenoportography was repeated to assess the extent of variceal obliteration. If collateral circulation to esophagogastric varices from the portal vein was detected, the aforementioned procedure was repeated until the blood flow in the varices ceased completely (Fig 1b). Finally, the portal pressure was measured again and the puncture tract was embolized with coils. Postinterventional supportive care included treatment for anemia and hypoproteinemia by blood transfusion, reduction of ascites with human serum albumin, and systemic prophylaxis with antibiotic agents for at least 7 days after the procedure. PSE was performed 7–10 days after PTVE in 31 patients in the PTVE/PSE group. The PTVE group was 34 patients who refused PSE after PTVE because they were worried about the postoperative complications of PSE. Briefly, splenic arterial angiography was performed by using a 5.0-F RH or Yashiro catheter (Terumo, Tokyo, Japan) to demonstrate the distribution of splenic arteries and collateral circulation routes. The tip of the catheter was placed at the splenic arteries in the middle or lower pole of the spleen. Polyvinyl alcohol particles (350–560 μm; Cook) were mixed with contrast media and 1 million units of penicillin G and/or 80 mg gentamicin. Under fluoroscopic guidance, the mixture was carefully injected manually through the catheter into the splenic arteries. PSE was performed progressively by means of repeated injections of polyvinyl alcohol under
1860
’
Variceal/Splenic Embolization for Esophagogastric Hemorrhage
Duan et al
’
JVIR
Figure 1. Images from a 59-year-old woman who presented to the emergency department with hematemesis and melena and was treated with PTVE/PSE. (a) CT angiography before PTVE revealed an enlarged gastric varix (arrow), dilated tortuous gastroesophageal varices, and a huge spleen as a result of hypersplenism (asterisk). (b) After occlusion of the enlarged gastric vein (arrow), which was opacified with contrast medium, splenoportography revealed a fine posterior gastric vein feeding the gastroesophageal varices by the collateral circulation. After embolization of this vein with two microcoils, splenoportography revealed complete occlusion of the varices. (c) After PTVE, with 8 days of supportive care, the patient was in hemodynamically stable condition. PSE was performed with the tip of the catheter placed at the splenic arteries in the middle pole of the spleen. (d) Two weeks after the procedure, contrast-enhanced CT showed the splenic infarction ratio to be controlled to approximately 50% of the original splenic volume.
angiography. The splenic infarction ratio was limited to approximately 50%–70% of the original splenic volume (Fig 1c) (15) by two interventional radiologists with 15 and 25 years of experience, who participated in all cases. The exact infarction rate was calculated from a CT scan (Fig 1d) obtained 2 weeks after PSE on a
3.1 Workstation (GE Medical Systems, Milwaukee, Wisconsin) with the use of volumetric analysis software as reported by Zhu et al (15). Supportive care and antipyretic agents for the treatment of postembolization syndrome were administered until the patients recovered and were discharged from the hospital.
Volume 25
’
Number 12
’
December
’
2014
Follow-up Study Sixty-five patients’ vital signs, complications during the procedures, postoperative complications, and the number of days of hospitalization required after PTVE were recorded. Ascites was monitored by US and hypertensive gastropathy by gastroscopy 5–7 days after PTVE or PTVE/PSE. All patients received US; three patients with HE in the PTVE group and two patients with labile hemodynamic findings did not undergo gastroscopy. The patients were followed up in the outpatient clinic at 1 week and 6, 12, and 24 months. Data were collected until recurrent bleeding, death, loss to follow-up, or completion of the study occurred; time to recurrent bleeding and cause of death were recorded. Recurrent bleeding was defined as the presence of “coffee-ground” vomitus, hematemesis, or melena and blood pressure lower than 90 mm Hg after a 24-h period of stable vital signs and hemoglobin levels after PTVE or PSE. Time to recurrence was defined as the time elapsed from the eradication of the acute hemorrhage to recurrent bleeding. Thrombocytopenia, white blood cell (WBC) count, red blood cell (RBC) count, albumin level, hemoglobin level, and Child–Pugh score were determined before PTVE and at 1 week and 6, 12, and 24 months after PTVE alone or PTVE/PSE.
Statistical Analysis Data are shown as means ⫾ standard deviation. Statistical analysis was performed by using SPSS software (version 13.0; IBM, Armonk, New York). The paired Student t test was used to determine statistically significant differences between two groups. One-way analysis of variance was used to compare WBC, RBC, and platelet counts and hemoglobin levels within groups over time. Significance was set at P o .05.
RESULTS Patient Characteristics The characteristics of the patients in the PTVE/PSE and PTVE-alone groups are shown in Table 1. The presence of ascites was determined by review of imaging before treatment. There were no significant differences in any characteristics between the two groups. The main clinical manifestations of hemorrhage were hematemesis and/ or black stools in all cases, and there were early signs of hemorrhagic shock in 44 cases. The amount of bleeding ranged from 1,000 mL to 1,900 mL, with an average of 1,400 mL. Preoperatively, the total transfusion volume of RBCs and plasma was 1,100–1,700 mL, with an average of 1,550 mL.
Outcomes and Complications of the Procedure The technical success rate of PTVE was 100% (N ¼ 65), with 104 varices successfully embolized in 65 patients.
1861
Table 1 . Characteristics of Study Patients Characteristic
PTVE/PSE (n ¼ 31)
PTVE (n ¼ 34)
Sex Male
21
23
Female Age (y)
10 40.1 ⫾ 10.9
11 39.7 ⫾ 9.9
HBV HCV
25 4
26 5
Budd–Chiari syndrome
2
2
0 7.18 ⫾ 1.72
1 7.09 ⫾ 2.11
4 21
6 23
Etiology (n)
Other Child–Pugh score Child–Pugh class A B
6
5
MELD score Ascites (n)
C
25.9 ⫾ 8.4 7
27.2 ⫾ 7.3 9
Follow-up (mo)
32.3 ⫾ 6.7
31.6 ⫾ 5.8
There were no significant differences between groups in any category. Values presented as means ⫾ standard deviation where appropriate. MELD scores ranges from 6 to 40, with higher scores indicating more severe disease. HBV ¼ hepatitis B virus, HCV ¼ hepatitis C virus, MELD ¼ Model for End-Stage Liver Disease, PSE ¼ partial splenic embolization, PTVE ¼ percutaneous transhepatic variceal embolization.
Active bleeding was controlled in all patients, with a 100% hemostasis rate. No varices ruptured during the procedure in the study. Portal vein pressures were measured before and after PTVE in all patients and found to differ significantly, from 22.9 mm Hg ⫾ 4.3 before the procedure to 29.1 mm Hg ⫾ 2.7 after the procedure (P o .001). No cases of pneumothorax, hemopneumothorax, or abdominal bleeding occurred during the procedure. One patient with poorly controlled hemorrhagic shock had a technical complication. In the process of gastric coronary vein embolization, the last coil tail extended to the portal vein as a result of the patient’s irritability for hemorrhagic shock with labile hemodynamic findings. During follow-up, no portal vein thrombosis developed after 5 days of anticoagulant therapy. The splenic infarction ratio was controlled to 60.5% ⫾ 8.7 of the original splenic volume in the PTVE/PSE group.
Side Effects and Complications after the Procedure Complications after treatment are shown in Table 2. There was no significant difference in the incidences of postoperative complications between the two groups with reference to Society of Interventional Radiology classification of complications. There were more cases of fever and abdominal pain in the PTVE/PSE group than in the PTVE group (P ¼ .035, P ¼ .038) and more cases of ascites and portal hypertensive gastropathy in the
1862
’
Variceal/Splenic Embolization for Esophagogastric Hemorrhage
PTVE group than in the PTVE/PSE group (P ¼ .003, P ¼ .014). There was no significant difference in the number of cases of HE or portal vein thrombosis. One case of portal vein thrombosis was found in each group by US 4 days after the procedure and resolved after 5 days of anticoagulant therapy. Two patients experienced HE in the PTVE group; one recovered and the other died 28 days after the procedure after restoration of liver function and alleviation of HE. No patient had a splenic abscess, abdominal bleeding, pulmonary embolism, or any other serious complication of treatment.
Follow-up Follow-up and prognosis are shown in Table 3. In the PTVE/PSE group, 31 patients were followed up at 1 week and 6 months and 30 patients at 12 and 24 Table 2 . Complications after Treatment
Complication
PTVE/PSE PTVE (n ¼ 31) (n ¼ 34) P Value
Fever Abdominal pain
25 23
11 10
.035 .038
Ascites
3
20
.003
Portal hypertensive gastropathy Hepatic encephalopathy
2 0
13 2
.014 NS
Portal vein thrombosis
1
1
NS
NS ¼ not significant, PSE ¼ partial splenic embolization, PTVE ¼ percutaneous transhepatic variceal embolization.
Duan et al
’
JVIR
months; four patients experienced recurrent bleeding during the follow-up period and one died of progressive liver failure 8 months after the procedure. In the PTVE group, 34 patients were followed up at 1 week, 31 at 6 months, 30 at 12 months, and 28 at 24 months; 15 patients experienced recurrent bleeding, including two who died of uncontrolled recurrent bleeding, three who died of progressive liver failure, and one who died of peritonitis. During follow-up, 17 of the 65 patients had variceal recurrent bleeding. CT angiography was performed to determine whether new collateral circulation to esophagogastric varices had been established. If new gastric coronary veins were established, PTVE was undertaken to control recurrent bleeding. If no gastric coronary veins were found, EVL, TIPS creation, or surgery was undertaken. Of the 17 patients with variceal recurrent bleeding, three in the PTVE/PSE group underwent PTVE again and one underwent EVL; in the PTVE group, 10 patients underwent PTVE, two underwent TIPS creation, and one underwent surgery. Of these 17 patients, none died or experienced recurrent bleeding during follow-up. No significant differences between groups were found in the number of hospitalization days after PTVE or the number of patients with complications or who died during follow-up. The time to recurrence in the PTVE/ PSE group was significantly longer than that in the PTVE group (P o .001). The cumulative recurrent bleeding rates at 6, 12, and 24 months were 3.2% (one
Table 3 . Follow-up and Prognosis Characteristic
PTVE/PSE (n ¼ 31)
PTVE (n ¼ 34)
P Value
Hospitalization days after PTVE (d) Recurrence
12.5 ⫾ 4.4 4
11.2 ⫾ 3.4 15
NS .038
Time to recurrence (mo)
9.5 ⫾ 3.9
5.2 ⫾ 2.5
o .001
3.2
20.6
o .001
6.7
36.7
.001
13.3
53.6
.004
Cumulative recurrent bleeding (%) 6 mo 12 mo 24 mo Child–Pugh score 1 wk
10.48 ⫾ 2.05
11.06 ⫾ 1.99
NS
6 mo 12 mo
9.12 ⫾ 1.02 8.19 ⫾ 1.98
9.87 ⫾ 2.16 9.32 ⫾ 2.03
.041 .036
24 mo
7.21 ⫾ 2.37
8.83 ⫾ 2.42
.012
1
6
NS
Death during follow-up Survival (%) 6 mo
100
94.1
NS
12 mo 24 mo
96.8 96.8
88.2 82.4
NS NS
0 1
2 3
NS NS
Cause of death (n) Recurrent bleeding Progressive liver failure Sepsis
0
0
NS
Peritonitis
0
1
NS
Values presented as means ⫾ standard deviation where appropriate. PSE ¼ partial splenic embolization, PTVE ¼ percutaneous transhepatic variceal embolization, NS ¼ not significant.
Volume 25
’
Number 12
’
December
’
2014
1863
of 31), 6.7% (two of 30), and 13.3% (four of 30) in the PTVE/PSE group and 20.6% (seven of 34), 36.7% (11 of 30), and 53.6% (15 of 28) in the PTVE group, respectively. There were significant differences in the cumulative recurrent bleeding rates between the PTVE/PSE and PTVE groups at 6 months (P o .001), 12 months (P o .001), and 24 months (P ¼ .004; Table 3). Survival rates at 6, 12, and 24 months were 100%, 96.8%, and 96.8% in the PTVE/PSE group and 94.1%, 88.2%, and 82.4% in the PTVE group, respectively. However, there were no statistically significant differences in these survival rates between the two groups. Child–Pugh score worsened from 7.18 ⫾ 1.72 to 10.48 ⫾ 2.05 in the PTVE/PSE group and from 7.48 ⫾ 2.05 to 11.06 ⫾ 1.99 in the PTVE group at 1 week after treatment. The scores in the PTVE/PSE group then improved to 9.12 ⫾ 1.02 at 6 months, 8.19 ⫾ 1.98 at 12 months, and 7.21 ⫾ 2.37 at 24 months, all of which were significantly different from those in the PTVE group at the respective time points (P o .05; Table 3).
Chronologic Changes in Peripheral Blood Cell Counts and Hemoglobin Levels Follow-up data on peripheral blood cell counts and hemoglobin levels at 1 week and 6, 12, and 24 months
after treatment are shown in Tables 4 and 5. Before treatment, there was no significant difference in peripheral blood cell counts or hemoglobin levels between the two groups (P 4 .05). After treatment, in the PTVE group, there were no significant differences in peripheral blood cell counts, platelet count, or hemoglobin level over time. In the PTVE/PSE group, WBC and platelet counts were lowest before treatment, reached a peak at 1 week, decreased during the 2-year follow-up period, and were significantly higher (P o .01) than those in the PTVE group at 1 week and 6, 12, and 24 months after treatment; RBC count and hemoglobin level increased significantly at 1 week, reached a peak at 6 months, decreased at 12 and 24 months after treatment, and were significantly higher (P o .01) than in the PTVE group at these time points.
DISCUSSION Although PTVE, endoscopy, TIPS creation, and pericardial devascularization have been used for the treatment of acute variceal hemorrhage for many years, each modality has drawbacks (4,16–18). Smith-Laing et al (19) showed a reduction or delay in recurrent bleeding after percutaneous transhepatic embolotherapy of
Table 4 . Changes in Platelet and WBC Counts Platelets (109/L)
WBCs (109/L)
Time
PTVE/PSE
PTVE
Pretreatment
40.5 ⫾ 11.4 (n ¼ 31)*
39.3 ⫾ 10.4 (n ¼ 34)
2.3 ⫾ 1.2 (n ¼ 31)*
PTVE/PSE
PTVE 2.2 ⫾ 0.9 (n ¼ 34)
42.1 ⫾ 14.8 (n ¼ 34)†
8.5 ⫾ 1.8 (n ¼ 31)
2.3 ⫾ 0.8 (n ¼ 34)†
6 mo 12 mo
93.5 ⫾ 13.5 (n ¼ 31)* 81.9 ⫾ 6.8 (n ¼ 30)*‡
41.9 ⫾ 17.2 (n ¼ 31)† 40.3 ⫾ 13.5 (n ¼ 30)†
6.2 ⫾ 0.9 (n ¼ 31)* 5.3 ⫾ 1.1 (n ¼ 30)*‡
2.5 ⫾ 1.2 (n ¼ 31)† 2.4 ⫾ 0.7 (n ¼ 30)†
24 mo
67.4 ⫾ 12.3 (n ¼ 30)*‡§
40.6 ⫾ 11.7 (n ¼ 28)†
4.3 ⫾ 1.7 (n ¼ 30)*‡§
2.5 ⫾ 1.3 (n ¼ 28)†
1 wk
112.5 ⫾ 14.8 (n ¼ 31)
PSE ¼ partial splenic embolization, PTVE ¼ percutaneous transhepatic variceal embolization, WBC ¼ white blood cell. *P o .01: platelet and WBC counts at 1 wk vs preprocedure and at 6, 12, and 24 mo. † P o .01: platelet and WBC counts in the PTVE/PSE group vs PTVE group. ‡ P o .01: platelet and WBC counts at 6 mo vs 12 mo and 24 mo. § P o .01: platelet and WBC counts at 12 mo vs 24 mo. Table 5 . Changes in RBC Count and Hemoglobin Level RBC (1012/L)
Hb (g/L)
Time
PTVE/PSE
PTVE
PTVE/PSE
PTVE
Preprocedure
2.5 ⫾ 0.4 (n ¼ 31)
2.4 ⫾ 0.5 (n ¼ 34)
72.3 ⫾ 19.2 (n ¼ 31)
71.2 ⫾ 19.4 (n ¼ 34)
1 wk 6 mo
2.6 ⫾ 0.3 (n ¼ 31)* 3.8 ⫾ 0.5 (n ¼ 31)*
2.5 ⫾ 0.7 (n ¼ 34)† 2.6 ⫾ 0.9 (n ¼ 31)†
81.5 ⫾ 11.7 (n ¼ 31)* 97.2 ⫾ 13.6 (n ¼ 31)*
75.3 ⫾ 13.7 (n ¼ 34)† 82.3 ⫾ 19.4 (n ¼ 31)†
12 mo
3.7 ⫾ 0.6 (n ¼ 30)*
2.4 ⫾ 0.8 (n ¼ 30)†
93.5 ⫾ 15.7 (n ¼ 30)*
75.2 ⫾ 16.7 (n ¼ 30)†
24 mo
3.2 ⫾ 0.7 (n ¼ 30)
83.3 ⫾ 14.7 (n ¼ 30)
76.3 ⫾ 15.5 (n ¼ 28)†
*
†
2.7 ⫾ 1.0 (n ¼ 28)
*
PSE ¼ partial splenic embolization, PTVE ¼ percutaneous transhepatic variceal embolization, RBC ¼ red blood cell. *P o .01: in the PTVE/PSE group, RBC count and Hb level preprocedure were significantly lower than those at different time points after treatment. RBC count and hemoglobin level were increased significantly at 1 wk, reached a peak at 6 mo, and declined at 12 and 24 mo after treatment. † P o .01: in the PTVE/PSE group, RBC count and hemoglobin level were significantly higher than in the PTVE group at different time points after treatment.
1864
’
Variceal/Splenic Embolization for Esophagogastric Hemorrhage
gastroesophageal varices. However, 65% of patients experienced recurrent bleeding at a mean of 4.6 months after successful PTVE in that study (19). Treatment with PTVE was first introduced by Lunderquist and Vang (10), but did not achieve widespread clinical acceptance because of the associated higher recurrent bleeding rate and the development of endoscopic treatment (16). However, the long-term recurrent bleeding rate after EVL remains high (4). Bian et al (4) reported that EVL performed to prevent variceal bleeding was associated with a recurrent bleeding rate of almost 30% and a 9%– 19% early recurrent bleeding rate. Variceal embolization by PTVE and variceal ligation by endoscopy are techniques that occlude portosystemic shunts to control bleeding (4,10,11). Gastrorenal shunts and gastric coronary veins are spontaneous portosystemic shunts formed to reduce portal hypertension; however, portal hypertension inevitably deteriorates after these shunts are occluded. Because cirrhosis cannot be effectively controlled, new gastrorenal shunts and gastric coronary veins are then established as a result of worsening portal hypertension (9). Endoscopic therapy is successful in most patients with controlled bleeding; however, in those who continue to experience bleeding and in whom the endoscopic view is obscured as a result of massive bleeding, TIPS creation is highly effective in achieving hemostasis (8,20). In view of the higher mortality rate associated with pericardial devascularization in the acute setting, TIPS creation is the favored rescue procedure for acute uncontrolled variceal hemorrhage (20). Emergent TIPS creation is recommended for the treatment of acute variceal bleeding in patients with moderate or severe liver failure and severe portal hypertension (2,8). Tesdal et al (9) reported cumulative survival rates of 75%, 69%, and 60% in 53 patients at 12, 24, and 48 months, respectively, after TIPS creation and embolotherapy. High Model for End-Stage Liver Disease scores may be associated with the high mortality rate associated with TIPS creation in the study of Tesdal et al (9). The cumulative recurrent bleeding rates in these 53 patients were 16% and 19% at 2 and 4 years after treatment, respectively. However, development of EVL and the implementation of emergent TIPS creation for acute variceal hemorrhage have been limited in China out of economic concerns and the relatively greater hepatic insufficiency in such patients in this country. With advances in embolization materials such as coils and microcoils, application of microcatheters, and improvement of interventional technologies, the high recurrent bleeding rate was decreased by means of PTVE. The cumulative recurrent bleeding rates at 6, 12, and 24 months after PTVE alone in the present study were 20.6%, 36.7%, and 53.6%, respectively, which are much better than those previously reported (19,21). However, in all patients, portal hypertension increased significantly from 22.88 mm Hg ⫾ 4.3 before PTVE to 29.1 mm Hg ⫾ 2.7 after PTVE. Because the deteriorating portal
Duan et al
’
JVIR
hypertension could not be effectively controlled, the number of patients with ascites or portal hypertensive gastropathy after PTVE was greater than that after PTVE/PSE. Apart from the reduction of splenic blood flow (14) and treatment of pancytopenia (15), PSE may be considered for the reduction of portal pressure in patients with hypersplenism (1) and therefore plays an important role in multiple antihemorrhage measures. Approximately 60%–70% of portal venous blood comes from the splenic vein in patients with portal hypertension. PSE decreases reflux in the splenic vein; hence, blood flow in the main portal vein is decreased, as is its pressure (14). Xu et al (14) reported that EVL with PSE decreased blood flow and pressure in the main portal vein to an extent equivalent to the combined effects of splenectomy and devascularization. Alzen et al (22) reported that portal vein pressure was reduced from 28.7 mm Hg ⫾ 7.4 to 24.3 mm Hg ⫾ 5.9 after PSE, representing 30–60% of the splenic volume. In the present study, portal hypertension was significantly increased after PTVE. However, the PTVE/PSE group had statistically significantly lower cumulative recurrent bleeding rates than the PTVE group at 6, 12, and 24 months and a significantly longer time to recurrence. Although changes in portal pressure after PTVE/ PSE were not measured during follow-up, our findings suggest that performing PSE 7–10 days after PTVE can reduce the increased portal pressure caused by PTVE. As a result of postembolization syndrome following PSE, there were more cases of fever and abdominal pain in the PTVE/PSE group than in the PTVE group. These adverse effects were temporary and can mostly be managed conservatively. However, the adverse effects of PTVE, such as ascites and portal hypertensive gastropathy, need to be controlled. Although PTVE was performed 7–10 days earlier than PTVE/PSE, there was no significant difference between the two groups in terms of the number of hospitalization days required after PTVE. Moreover, PSE is much less invasive than traditional splenectomy and can help avoid the risks of surgery intra- and postoperatively (22). There are many advantages concerning PSE, the most important probably being the preservation of the immune function of the spleen and the treatment of leukocytopenia and thrombocytopenia caused by hypersplenism. The results of the present study are in accordance with the peripheral blood cell and hemoglobin level change trends reported by Zhu et al (15) in alleviating hypersplenism. Hu et al (23) achieved cumulative recurrent bleeding rates of 6.6%, 27.7%, 44.1%, and 65.6% at 1, 3, 5, and 8 years after PTVE combined with PSE. With the same treatment, Liu et al (24) achieved cumulative recurrent bleeding rates of 12.5% (six of 47), 24.5% (11 of 45), and 27.9% (12 of 43) at 12, 24, and 36 months, respectively. As a result of poor hepatic function and adverse reactions to PTVE, the patients in the present study
Volume 25
’
Number 12
’
December
’
2014
would not have tolerated the postembolization syndrome caused by concurrent PSE and PTVE, as described by Hu et al (23) and Liu et al (24). Instead, in the PTVE/PSE group, PTVE was followed by PSE after 7–10 days, when the patients were in hemodynamically stable condition and their liver function had improved with supportive care. Although the percentages of our patients with Child–Pugh class C disease and hemodynamic instability before treatment were more than those reported in the studies of Hu et al (23) and Liu et al (24), cumulative recurrent bleeding rates of 3.2%, 6.7%, and 12.3% were achieved at 6, 12, and 24 months in our PTVE/PSE group. We also found that hepatic functional reserve was maintained during follow-up in patients treated with PTVE/PSE compared with PTVE alone. The present study has several limitations. First, the study was neither blinded nor randomized; therefore, the data could have been affected by various potential biases. However, these may have been partially overcome by the independent follow-up and the similar size of the two groups. The procedures studied here are complicated to perform, and the optimal interval between PSE and PTVE is unclear, which might limit the use of this combined method as the first-choice treatment for acute variceal massive bleeding. In addition, the study was retrospective, with a limited number of cases. A prospective randomized controlled study with a large number of subjects is needed to investigate whether PTVE/PSE is a superior treatment for patients with acute variceal massive bleeding with portal hypertension and cirrhosis. In conclusion, PTVE/PSE is a relatively safe and effective method for the treatment of acute massive bleeding from esophagogastric varices in patients with cirrhosis and portal hypertension who cannot be considered for other interventions. PTVE/PSE not only has longterm efficacy in alleviating hypersplenism, but also treats acute variceal massive bleeding and prevents recurrent bleeding. PTVE/PSE should also help maintain hepatic reserve and may improve the prognosis of cirrhosis in these patients.
REFERENCES 1. Kirby JM, Cho KJ, Midia M. Image-guided intervention in management of complications of portal hypertension: more than TIPS for success. Radiographics 2013; 33:1473–1496. 2. Garcia-Pagan JC, Caca K, Bureau C, et al. Early use of TIPS in patients with cirrhosis and variceal bleeding. N Engl J Med 2010; 362:2370–2379. 3. Garcia-Tsao G, Sanyal AJ, Grace ND, Carey W. Prevention and management of gastroesophageal varices and variceal hemorrhage in cirrhosis. Hepatology 2007; 46:922–938.
1865
4. Bian S, Tian XG, Hu JH, Wang GC, Zhang CQ. Percutaneous transhepatic variceal embolization combined with endoscopic ligation for the prevention of variceal rebleeding. J. Dig Dis 2013; 14:388–395. 5. Okada M, Nakashima Y, Kishi T, et al. Percutaneous transhepatic obliteration for massive variceal rectal bleeding. Emerg Radiol 2012; 19: 355–358. 6. Palejwala AA, Smart HL, Hughes M. Multiple pulmonary glue emboli following gastric variceal obliteration. Endoscopy 2000; 32(suppl):S1–S2. 7. Lo GH, Lai KH, Cheng JS, Huang RL, Wang SJ, Chiang HT. Prevalence of paraesophageal varices and gastric varices in patients achieving variceal obliteration by banding ligation and by injection sclerotherapy. Gastrointest Endosc 1999; 49:428–436. 8. Loffroy R, Estivalet L, Cherblanc V, et al. Transjugular intrahepatic portosystemic shunt for the management of acute variceal hemorrhage. World J Gastroenterol 2013; 19:6131–6143. 9. Tesdal IK, Filser T, Weiss C, Holm E, Dueber C, Jaschke W. Transjugular intrahepatic portosystemic shunts: adjunctive embolotherapy of gastroesophageal collateral vessels in the prevention of variceal rebleeding. Radiology 2005; 236:360–367. 10. Lunderquist A, Vang J. Sclerosing injection of esophageal varices through transhepatic selective catheterization of the gastric coronary vein. A preliminary report. Acta Radiol Diagn (Stockh) 1974; 15:546–550. 11. Pereiras R, Viamonte M, Jr., Russell E, LePage J, White P, Hutson D. New techniques for interruption of gastroesophageal venous blood flow. Radiology 1977? 124:313–323. 12. Koconis KG, Singh H, Soares G. Partial splenic embolization in the treatment of patients with portal hypertension: a review of the English language literature. J Vasc Interv Radiol 2007; 18:463–481. 13. Romano M, Giojelli A, Capuano G, Pomponi D, Salvatore M. Partial splenic embolization in patients with idiopathic portal hypertension. Eur J Radiol 2004; 49:268–273. 14. Xu RY, Liu B, Lin N. Therapeutic effects of endoscopic variceal ligation combined with partial splenic embolization for portal hypertension. World J Gastroenterol 2004; 10:1072–1074. 15. Zhu K, Meng X, Qian J, et al. Partial splenic embolization for hypersplenism in cirrhosis: a long-term outcome in 62 patients. Dig Liver Dis 2009; 41:411–416. 16. Rosch J, Keller FS. Transjugular intrahepatic portosystemic shunt: present status, comparison with endoscopic therapy and shunt surgery, and future perspectives. World J Surg 2001; 25:337–345. 17. Tian X, Wang Q, Zhang C, Liu F, Cui Y, Liu J. Modified percutaneous transhepatic variceal embolization with 2-octylcyanoacrylate for bleeding gastric varices: long-term follow-up outcomes. AJR Am J Roentgenol 2011; 197:502–509. 18. Lee JH, Han HS, Kim HA, Koo MY. Long-term results of fundectomy and periesophagogastric devascularization in patients with gastric fundal variceal bleeding. World J Surg 2009; 33:2144–2149. 19. Smith-Laing G, Scott J, Long RG, Dick R, Sherlock S. Role of percutaneous transhepatic obliteration of varices in the management of hemorrhage from gastroesophageal varices. Gastroenterology 1981; 80: 1031–1036. 20. McAvoy NC, Hayes PC. The use of transjugular intrahepatic portosystemic stent shunt in the management of acute oesophageal variceal haemorrhage. Eur J Gastroenterol Hepatol 2006; 18:1135–1141. 21. Sos TA. Transhepatic portal venous embolization of varices: pros and cons. Radiology 1983; 148:569–570. 22. Alzen G, Basedow J, Luedemann M, Berthold LD, Zimmer KP. Partial splenic embolization as an alternative to splenectomy in hypersplenism– single center experience in 16 years. Klin Padiatr 2010; 222:368–373. 23. Hu YB, Feng GS, Liang HM, et al. Long term effects of interventional devascularization in treating cirrhosis patients with portal hypertension. Chin J Radiol 2009; 43:853–857. 24. Liu W, Chen GS, Sun HL, et al. Comparison of interventional devascularization and surgical devascularization in treatment of upper gastrointestinal hemorrhage for portal hypertensive. Chin J Radiol 2009; 43: 748–752.
