| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Samsung Medical Center, Sungkyunkwan University
School of Medicine, Ilwon-dong 50, Kangnam-ku, Seoul 135-710, Korea.
2 Department of Medicine, Division of Gastroenterology, Samsung Medical Center,
Sungkyunkwan University School of Medicine, Seoul, Korea.
revised March 17, 2007;
accepted after revision June 25, 2007.
Address correspondence to S. W. Shin
(swshin{at}smc.samsung.co.kr)
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. From November 2002 through October 2005, 49 consecutive patients with gastric varices were treated by balloon-occluded retrograde transvenous obliteration. The sclerosant was injected through the outflow veins during balloon occlusion. Immediate postprocedural CT scans were obtained to evaluate the procedural details. Recurrence and rebleeding of gastric varices and worsening of esophageal varices were evaluated by endoscopic examination and CT. The survival rates and prognostic factors after the procedure were also assessed.
RESULTS. There were six procedural failures and two procedure-related deaths. Disappearance or marked shrinkage of the treated gastric varices with no recurrent gastric variceal bleeding was noted in 39 patients (79.6% clinical success rate). Approximately two thirds of our patients experienced worsening of esophageal varices during the median follow-up period of 457 days. The cumulative survival rates at 1 year and 3 years after balloon-occluded retrograde transvenous obliteration were 83.1% and 65.7%, respectively. The prognostic factors associated with survival were the preprocedural Child-Pugh classification and the total bilirubin level. The survival rates and procedural outcomes for the patients with severely compromised liver function were poor.
CONCLUSION. Balloon-occluded retrograde transvenous obliteration is an effective treatment for the obliteration of gastric varices. However, application of this procedure to severely compromised patients should be considered carefully.
Keywords: balloon-occluded retrograde transvenous obliteration • embolization • gastric varices • interventional radiology
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Gastric varices with spontaneous gastrosystemic shunts that may be catheterized have been recently treated by retrograde sclerotherapy through the shunts or the outflow veins of gastric varices, which is the so-called balloon-occluded retrograde transvenous obliteration of gastric varices [5, 16-22]. The bleeding control rate of gastric varices after balloon-occluded retrograde transvenous obliteration has been described as greater than 90%, and the procedure has been reported to have an important advantage over other treatment strategies because it can be attempted in patients with a poor hepatic functional reserve and even in patients with encephalopathy [5, 16, 17]. The procedure has been widely performed in Japan, but there are few reports of balloon-occluded retrograde transvenous obliteration from countries other than Japan. The purpose of this study is to describe the clinical outcomes of balloon-occluded retrograde transvenous obliteration and to detail its procedural aspects and complications based on our institutional experience.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All patients but two had experienced at least one episode of gastric variceal bleeding before balloon-occluded retrograde transvenous obliteration, as was confirmed by endoscopic examinations; and the period between balloon-occluded retrograde transvenous obliteration and the endoscopic examinations ranged from 0 to 28 days (mean, 5.4 days). In four patients, endoscopy revealed active (spurting) bleeding from gastric varices despite a trial of endoscopic hemostasis, and balloon-occluded retrograde transvenous obliteration was performed for these patients on an emergency basis after the insertion of Sengstaken-Blakemore tubes.
Endoscopy in 43 patients showed signs of recent gastric variceal bleeding, such as erosion or blood clots that adhered to the gastric varices, and these patients underwent elective balloon-occluded retrograde transvenous obliteration. The remaining two patients underwent balloon-occluded retrograde transvenous obliteration to prevent gastric variceal bleeding. The presence of a spontaneous gastrosystemic shunt was confirmed by performing contrast-enhanced CT before the balloon-occluded retrograde transvenous obliteration. Two patients showed bland portal vein thrombus (nontumorous) that partially blocked the main portal veins.
Balloon-Occluded Retrograde Transvenous Obliteration Technique
All patients gave informed consent before the procedures. According to the
diameter of the gastrosystemic shunts as measured on CT images, an occlusion
balloon catheter of an appropriate diameter (8.5-33 mm) was inserted in a
retrograde fashion into the gastrosystemic shunts; this was usually done via
the right femoral vein. The anatomic types of gastrosystemic shunts are
summarized in Table 1. To
facilitate catheterization of the gastrorenal shunt, the end portion of the
occlusion balloon catheter was steamed and formed a zigzag shape. After
insertion of an 8- to 12-French vascular sheath, the occlusion balloon
catheter was advanced over a 0.035-inch glidewire and was inserted into the
left adrenal vein (a portion of the gastrosystemic shunt) through the left
renal vein. After negotiation of the occlusion balloon catheter into the
shunt, the balloon was inflated and a retrograde venogram was obtained with
manual contrast injection through the occlusion balloon catheter. The
catheterized shunts were mostly single gastrorenal shunts (n = 45)
and, in these cases, we were usually able to achieve full opacification of all
gastric varices seen on the retrograde venogram.
|
This procedure was followed by retrograde injection of a sclerosing agent, which was a mixture of 5% ethanolamine oleate and iodized oil (Lipiodol, Andre Guerbet) or contrast medium (Ultravist 300 [iopromide], Schering [now Bayer Healthcare]) that was mixed in a ratio of 4:1-6:1. The sclerosant was slowly injected until the gastric varices became completely opacified and the feeding veins from the portal or splenic veins could be visualized. For patients with leaking collateral veins such as the inferior phrenic or paravertebral veins that prevented full opacification of gastric varices, we occluded these collateral veins using spring coils (Tornado, Cook) or gelatin sponge pledgets (Cutanplast, Mascia Brunelli) through a selectively catheterized 3-French microcatheter (Microferret, Cook) (Fig. 1A, 1B, 1C, 1D, 1E, 1F).
|
|
|
|
|
|
|
|
|
The balloon inflation time was decided according to the operator's discretion, but usually, the greater the size of the gastric varix, the longer the inflation time. Thereafter, as much of the sclerosant as possible was aspirated. If blood was freely aspirated with the residual sclerosant through the balloon catheter, then the catheter was kept inflated for a few additional hours and a few milliliters of supplemental sclerosant was injected because the free aspiration of blood meant inadequate thrombus formation in the gastric variceal complex (inflow veins, gastric varices, and outflow veins).
Finally, after confirmation that blood flow through the catheter had ceased, the balloon catheter was carefully deflated and removed under fluoroscopic observation while closely observing whether the formed thrombus remained stationary. Procedure times ranged from 30 to 120 minutes, with 5-20 minutes of fluoroscopy times, which depended largely on the complexity of the gastrosystemic shunts.
Follow-Up
As long as the patient's clinical condition permitted it, CT was routinely
obtained 2-4 days after the balloon-occluded retrograde transvenous
obliteration to evaluate the immediate procedural outcome. Thereafter,
follow-up CT and endoscopy were performed every 3-6 months. The follow-up
interval was decided at the discretion of the attending physician and
according to the patient's condition, which included such factors as
rebleeding, accompanying esophageal varices, or HCC. Recurrence and rebleeding
of gastric varices were evaluated by performing follow-up endoscopy or CT, and
follow-up evaluation of the esophageal varices was performed via endoscopy.
When the follow-up endoscopy revealed newly developed esophageal varices, red
spots on preexisting esophageal varices, or esophageal variceal bleeding, the
esophageal varices were regarded as having worsened, and endoscopic treatments
were then performed accordingly.
Definitions and Statistical Analysis
Procedure-related deaths were defined as deaths occurring within 24 hours
of the procedure. Procedural success was defined as successful injection of
the sclerosant with full opacification of the gastric varices and complete
thrombus formation in the gastric varices as confirmed on immediate follow-up
CT. Clinical success was defined as marked shrinkage or disappearance of
gastric varices, no recurrent gastric variceal bleeding, and no significant
procedural adverse effects. The values were expressed as means ± SDs,
medians, or percentages. The survival rates were calculated according to the
Kaplan-Meier method.
To evaluate the prognostic factors, the following parameters were included
in the multiple regression analysis using the Cox proportional hazard model:
the presence or absence of active gastric variceal bleeding at the time of the
procedure, procedural success versus failure, coexisting HCC, coexisting
esophageal varices, the Child-Pugh classification (A or B vs C), and the total
bilirubin level (< 3.5 vs 
3.5 mg/dL). Because we found that patients
with a high total bilirubin level had a poor prognosis after balloon-occluded
retrograde transvenous obliteration, the total bilirubin level was
independently analyzed.
In addition, we separately analyzed the normal (at least 1 month before the procedure) and the preprocedural (immediately before the procedure) values of the Child-Pugh classification and the total bilirubin level because acute or subacute variceal bleeding itself can affect the measurement of serum bilirubin and consequently the Child-Pugh classification. The results of this analysis are presented as relative risk estimates with corresponding 95% CIs. Values for p of less than 0.05 were considered to indicate significant differences. All computations were performed using statistical software (version 11.5, SPSS).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
There were six procedural failures. One was due to the complexity of the gastrosystemic shunts, which were composed of gastrorenal, gastrocaval, and gastropericardiophrenic shunts that prevented complete occlusion of the shunts; and two were due to incomplete thrombus formation in the gastric varices because the amount of the injected sclerosant was not enough for full opacification of gastric varices. The remaining three failures were due to large gastrorenal shunts that we could not occlude even with the use of our largest balloon catheter. Of these six procedural failures, one patient died of persistent variceal bleeding 5 days after the procedure, one patient was treated by endoscopic variceal ligation with injection sclerotherapy, one patient was discharged with hopelessly deteriorating hepatic function and the clinical suspicion of concomitant HCC rupture, and three patients showed spontaneous control of their bleeding with supportive care only. However, two patients among the three spontaneous responders died from rebleeding and hepatic failure, respectively, within 2 months after the balloon-occluded retrograde transvenous obliteration.
Except for these eight patients, the immediate postprocedural CT scans showed complete thrombus formation and no contrast enhancement of the gastric varices in all patients, so the procedural success rate was 83.7% (41/49).
Immediate postprocedural CT scans were obtained in 46 patients at 1-11 days (mean, 3.1 days) after balloon-occluded retrograde transvenous obliteration. On these CT scans, newly developed ascites or aggravated ascites was seen in 20 patients (43.5%), and new pleural effusion was noted in 33 patients (71.7%). The amounts of ascites and pleural effusion gradually decreased and returned to the baseline after a few months in most patients. In two patients (4.3%), main portal vein thrombosis was noted from the extension of the formed thrombus in the afferent gastric veins. One of these two patients, who had already had duodenal varices, experienced worsening and bleeding of his duodenal varices 6 months after the balloon-occluded retrograde transvenous obliteration, and his duodenal varices were treated with another obliteration procedure through the duodenogonadal shunt because endoscopic approaches had failed (Fig. 3A, 3B, 3C). Pulmonary thromboembolism developed in two patients (4.3%), and one of these patients complained of transient dyspnea. One patient (2.2%) showed left renal vein thrombosis that caused permanent renal vein occlusion, but this was clinically silent. On the other hand, the two patients who already had bland portal vein thrombus that partially blocked the main portal vein showed temporary resolution of the thrombus, although thrombosis recurred several months later.
|
|
|
Clinical Success and Follow-Up Results for the Varices
Among the procedural successes, one patient was discharged because of a
hopeless condition, and another died within 2 months after the procedure due
to deteriorating hepatic function. Both these patients had Child-Pugh
classifications of C and preprocedural total bilirubin levels of 13.1 and 15.4
mg/dL. After excluding these two patients, there was neither recurrence nor
rebleeding of the treated gastric varices during the follow-up period (median,
574 days; mean, 637 ± 324 days; range, 78-1,275 days). The follow-up
examinations revealed marked shrinkage or disappearance of the gastric
varices, yielding a clinical success rate of 79.6% (39/49). The clinical
success rate reached 95.1% once procedural success was achieved (39/41).
Follow-up endoscopy was performed in 31 patients. Of these patients, 21 patients (67.7%) experienced worsening of their esophageal varices during the follow-up period (median, 457 days; mean, 539 ± 329 days; range, 81-1,276 days), and bleeding from esophageal varices occurred in 10 patients (32.2%, 10/31). All of these patients were successfully treated endoscopically except for one patient who died from uncontrolled esophageal bleeding and subsequent multiorgan failure. On follow-up endoscopy, gastric cardiac varices developed in five patients (16.1%, 5/31); the varices were irrelevantly located from the fundal varices that were previously treated with balloon-occluded retrograde transvenous obliteration. All of these cardiac varices were noted in patients who displayed worsening of their esophageal varices.
Survival Rates and Prognostic Factors
The overall cumulative survival rates at 1, 2, and 3 years after
balloon-occluded retrograde transvenous obliteration were 83.1%, 75.9%, and
65.7%, respectively (Fig. 4A,
4B,
4C). According to the
preprocedural Child-Pugh classifications, the 3-year survival rates were
83.3%, 69.2%, and 34.8% for classifications A, B, and C, respectively. The 1-,
2-, and 3-year survival rates for the patients with a preprocedural total
bilirubin level of less than 3.5 mg/dL were 92.0%, 82.5%, and 73.3%; and the
1-, 2-, and 3-year survival rates for the patients with a preprocedural total
bilirubin level of 3.5 mg/dL or more were 41.1%, 20.6%, and 20.6%,
respectively.
|
|
|
Active gastric variceal bleeding, coexisting HCC, and esophageal varices had no significant impact on the overall survival after balloon-occluded retrograde transvenous obliteration. On multivariate analysis, a preprocedural Child-Pugh classification of C and a preprocedural total bilirubin level were the significant risk factors for overall survival (Fig. 4A, 4B, 4C and Table 2). Although procedural failure was not a significant independent factor for survival on multivariate analysis, a trend for a poor prognosis after procedural failure was noted.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Balloon-occluded retrograde transvenous obliteration has been introduced as a treatment method that aims to obliterate the gastric varix itself. Since Kanagawa et al. [20] reported their balloon-occluded retrograde transvenous obliteration results with using ethanolamine oleate as a sclerosant, studies—mainly from Japan—have shown that the procedure can successfully induce thrombosis in gastric varices, and these thrombosed varices show marked shrinking and complete resolution. The reported efficacy of balloon-occluded retrograde transvenous obliteration of gastric varices has been 87-100%, and the relapse rate is 0-10% [5, 16-18, 20-22], which shows excellent outcomes. Although the clinical success rate depends on the definition of clinical success and the method of evaluation, our results, with a clinical success rate of 79.6%, seem somewhat lower than previously reported results. However, once thrombus had been successfully formed in gastric varices, there was neither recurrence nor rebleeding from the treated gastric varices, in which we noted almost complete shrinkage or their disappearance (a 95.1% clinical success rate among the successful procedures). The two clinical failures among the procedural successes were not because of recurrence or rebleeding of gastric varices; rather, they were due to hepatic failure. In addition, although prophylactic cases constituted more than 50% of cases in previous studies in which patient characteristics were identified [22-24], only two patients (4%) in our study received prophylactic balloon-occluded retrograde transvenous obliteration.
Some authors have described their long-term follow-up results after balloon-occluded retrograde transvenous obliteration [21-25]. According to these results, the overall 3-year survival rates were 75-87.4%. The 3-year survival rate of our study was 71.9%, slightly lower than rates of previous studies. However, we think that the preprocedural liver function of our patient group was poorest when compared with the studies for which the preprocedural biochemical data were available [21-23]. The mean preprocedural Child-Pugh score and serum bilirubin level of patients from those previous studies were less than 7 and less than 2.0 mg/dL, respectively [21-23], whereas those values for our patients were 8.7 and 3.3 mg/dL, respectively. A few studies have recognized the preprocedural Child-Pugh classification to be a significant factor that affects patient survival [21, 23, 24]. The difference in survival rates between our study and the other studies may be partly explained by the difference in the preprocedural liver function between the studies. Similar to the previous studies, the preprocedural Child-Pugh classification was an important prognostic factor associated with survival in our study; and among the parameters that made up the Child-Pugh classification, the total bilirubin level was an independent prognostic factor.
As we have described, the efficacy of balloon-occluded retrograde transvenous obliteration for treating gastric varices has been described as excellent, and some authors have insisted that the procedure could be used as a prophylactic treatment technique before the rupture of gastric varices. Furthermore, this strategy may prolong patient survival [5, 22-24]. Chikamori et al. [22] even claimed that prophylactic balloon-occluded retrograde transvenous obliteration may contribute to improved long-term survival in the Child-Pugh classification C group. However, our study showed somewhat different results: The 3-year survival rate of the preprocedural Child-Pugh classification C group in our study was 34.8%, which is lower than those of other studies [21-23]. Especially for patients with a preprocedural total bilirubin level of 3.5 mg/dL or more, the 3-year survival rate was only 20.6%. Including two procedure-related deaths, we encountered eight patients who died or were discharged as hopeless because of their deteriorating hepatic failure within 2 months after their obliteration procedures; of these patients, seven had a preprocedural serum total bilirubin level equal to or greater than 3.5 mg/dL and a Child-Pugh classification of C or late B (score of 8 or 9). Conversely, nine patients in our study group had a preprocedural total bilirubin level of 3.5 mg/dL or more; and among these patients, 78% (7/9) had a dismal prognosis because of pending hepatic failure. These poor procedural outcomes might have affected the poor long-term survival rates in our study.
From these findings, we carefully conclude that it seems difficult for balloon-occluded retrograde transvenous obliteration to be universally applied as a prophylactic treatment for gastric varices in patients with cirrhosis; and even when it is used therapeutically, the benefits and risks should be weighed immediately before the obliteration procedure, especially for patients whose liver function appears to be profoundly compromised. Balloon-occluded retrograde transvenous obliteration has been thought to improve portal venous flow because it occludes the gastrosystemic shunt and diverts the blood flow through the shunt to the liver [16, 26, 27]. Although we cannot say the exact reason severely compromised patients show such poor short-term results after the procedure, we speculate that the vulnerable livers of severely cirrhotic patients cannot accommodate the suggested sudden portal hemodynamic change that occurs from balloon-occluded retrograde transvenous obliteration. Otherwise, the poor prognosis of these compromised patients might be just a reflection of their spontaneous disease progression, regardless of the results of balloon-occluded retrograde transvenous obliteration.
With respect to procedure-related death, we encountered one patient (who subsequently died) who had portal vein tumor thrombus before the balloon-occluded retrograde transvenous obliteration whose Child-Pugh classification was A. It is not certain whether portal vein thrombosis is a contraindication to balloon-occluded retrograde transvenous obliteration. We also observed the temporary resolution of bland portal vein thrombus after this procedure. However, caution should be exerted when performing the procedure in patients with a preexisting obstructive lesion in the portal vein because it is plausible that the obliteration procedure may acutely increase portal venous flow and portal pressure.
As we have described, balloon-occluded retrograde transvenous obliteration may augment the hepatopetal portal flow, which may lead to an increase of portal venous pressure and also to the unique complications of the obliteration procedure, such as aggravation of esophageal varices and an increased amount of ascites. The 3-year cumulative rate of worsening esophageal varices has been reported to be 24.9-58% [21-23]. In our study, 67.7% of patients experienced worsening of their esophageal varices during a median follow-up period of 457 days, which is a seemingly higher rate than those of previous studies. This higher rate may have been the result of the differences in our analysis method because we calculated the rate of esophageal variceal worsening for those patients who underwent endoscopic examinations and we excluded patients who had no long-term follow-up. Nonetheless, esophageal varices were aggravated in a substantial portion of our patients, so we believe patients who have undergone balloon-occluded retrograde transvenous obliteration should be carefully monitored with endoscopic examinations and treated accordingly.
As for ascites, 43.5% of the patients in our study experienced new or aggravated ascites on the basis of the immediate follow-up CT. The rate of worsening ascites has been reported to be 0-7% [19, 21, 22, 28], and our results seem to show a higher rate than those in previous studies. The mean time between balloon-occluded retrograde transvenous obliteration and the postprocedural CT in our study was 3.1 days, whereas the same interval was usually 1 week in the other studies. We think that this difference in the time interval may partially explain the difference in the rate of worsening ascites because the amount of ascites would decrease as time passes.
Ethanolamine oleate is the most commonly used sclerosant in balloon-occluded retrograde transvenous obliteration; its reported complications include severe renal dysfunction [29], pulmonary edema [30], disseminated intravascular coagulation [31], and anaphylactic reaction [5]. Of these complications, renal dysfunction is not rare because the free hemoglobin and hemoglobinuria induced by intravascular hemolysis from the use of ethanolamine oleate may cause renal tubular dysfunction [29]. To avoid renal dysfunction, haptoglobin has been prophylactically administered IV [5, 16-20, 23, 24, 27, 28] because haptoglobin conjugates hemoglobin and may prevent renal tubular dysfunction [29]. Some authors have limited the dose of ethanolamine oleate to 20-30 mL [5, 21, 22].
We did not prophylactically administer haptoglobin because it is not available in our country. The reported incidence of hemoglobinuria after balloon-occluded retrograde transvenous obliteration is 15-100% despite the use of haptoglobin [19, 21, 22, 28], and the incidence of hemoglobinuria in our study was 53.1%. Hemoglobinuria did not cause renal failure in our study except for one presumed case of postprocedural multiorgan failure. In addition, we sometimes used 40-45 mL of ethanolamine oleate for complete filling of the gastric varices. Because our two procedural failures resulted from the use of an insufficient amount of ethanolamine oleate, we believe that for achieving technically successful balloon-occluded retrograde transvenous obliteration, the amount of ethanolamine oleate should be the amount needed for complete filling of all the varices. In addition, the systemic effects of ethanolamine oleate will be alleviated because outflow veins are being occluded during the obliteration procedure.
One advantage of balloon-occluded retrograde transvenous obliteration is that it is technically easier and safer to perform than the other percutaneous procedure, TIPS. However, in our experience the procedural failure usually occurred during the early period after we started to perform the obliteration. This means that performing the obliteration procedure has its own learning curve, and interventional radiologists should try to decrease procedural failures, especially in patients with complex gastrosystemic shunts. Although the gastrorenal shunt is the most common access through which balloon-occluded retrograde transvenous obliteration is performed, occlusion of the gastrorenal shunt alone is sometimes inadequate for successful obliteration in patients with complex shunts. In these cases, occlusion of all the shunts is needed. Careful inspection of the preprocedural CT scans makes it possible to predict this kind of complex shunt.
We encountered a few cases of major venous thrombosis after balloon-occluded retrograde transvenous obliteration that were caused by the extension of the formed thrombus into the variceal complex. To prevent this complication, the proper positioning of the balloon catheters as deep into the outflow veins as possible is needed, and excessive thrombus formation in the afferent veins of gastric varices should be avoided.
In conclusion, once thrombus was completely formed in gastric varices after balloon-occluded retrograde transvenous obliteration, as determined on the immediate postprocedural CT scans, this procedure seems to effectively obliterate gastric varices. It provides far lower rebleeding and recurrence rates than other methods. However, because patients whose liver function is already profoundly compromised may have a poor prognosis after the obliteration procedure, the benefits and risks from this procedure should be weighed for these patients.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
