A comparison of the results of arterial embolization for ...

AR-11-0344

Original article

A comparison of the results of arterial embolization for bleeding and non-bleeding gastroduodenal ulcers Romaric Loffroy1, MingDe Lin2, Carol Thompson3, Amith Harsha4 and Pramod Rao1 1

Russell H Morgan Department of Radiology and Radiological Science, Division of Vascular and Interventional Radiology, Johns Hopkins Hospital, Baltimore, Maryland; 2Clinical Informatics, Interventional, and Translational Solutions (CIITS), Philips Research North America, Briarcliff Manor, New York; 3Johns Hopkins Biostatistics Center, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland; 4 Department of Neuroradiology and Biomedical Engineering, Johns Hopkins Medical Institutions, Baltimore, Maryland, USA Correspondence to: Romaric Loffroy. Email: [email protected]

Abstract Background: Although some authors have advocated the practice of arterial embolization for angiographically negative acute hemorrhage from gastroduodenal ulcers, this technique remains controversial. Purpose: To compare the results of arterial embolization for bleeding (BU) and non-bleeding (NBU) gastroduodenal ulcers. Material and methods: Transcatheter embolization was performed in 57 patients (39 men, 18 women, mean age 69.8 years) who experienced acute bleeding from gastroduodenal ulcers. At the time of embolization active contrast extravasation was seen in 36 of 57 patients, while in the remaining 21 patients embolization was based on endoscopic findings. Patient demographics, clinical success, need for re-intervention secondary to re-bleeding, and 30-day complication and mortality rates were reviewed and compared between the two groups by using statistical analyses. Results: In the BU group, the gastroduodenal artery (GDA) was embolized in 31 patients (86.1%), the left gastric artery (LGA) in three patients (8.3%), and the left gastroepiploic artery (LGEA) in two patients (5.6%). In the NBU group, the GDA was embolized in 18 patients (85.7%), and the LGA in three patients (14.3%). Clinical success (61.9 vs. 75.0%, P ¼ 0.30), need for re-intervention (38.1 vs. 27.8%, P ¼ 0.42), and 30-day complication (9.5 vs. 5.6%, P ¼ 0.57), and mortality (28.6 vs. 25%, P ¼ 0.77) rates were not statistically different between the two groups. Embolization in patients in NBU group did not have impact on clinical success (OR, 0.54; 95%CI, 0.17–1.72; P ¼ 0.30). Conclusion: Arterial embolization in patients with angiographically NBU is as safe and effective as embolization in patients with BU.

Keywords: Upper gastrointestinal tract, acute hemorrhage, peptic ulcers, endoscopy, embolization Submitted May 8, 2011; accepted for publication August 22, 2011

Acute bleeding is the leading complication of peptic ulcer disease, which contributes to approximately half of the cases of upper gastrointestinal bleeding (1, 2). Bleeding from peptic ulcers is fatal in 5 – 10% of patients, a proportion that has not changed substantially over the past two decades (3, 4). Emergent endoscopy remains the primary investigation in upper gastrointestinal bleeding, allowing for identification and localization of the bleeding source (5 – 7). Bleeding stops spontaneously in 80% of cases but recurs or persists in 20%. Severe bleeding despite conservative medical treatment or endoscopic intervention occurs in

5% of patients, requiring surgery or transcatheter arterial embolization (5). In many institutions, endovascular management is now considered to be the first-line treatment for massive bleeding from gastroduodenal ulcers refractory to endoscopic treatment (8 – 15). Massive bleeding is often intermittent in nature. Contrast medium extravasation from a bleeding artery is frequently not identified using angiography, leading to difficulties in finding the exact site for embolization (16). In this situation, the likely offending artery may be embolized on the basis of endoscopic findings in the hope of preventing further hemorrhage. Acta Radiologica 2011; 00: 1– 7. DOI: 10.1258/ar.2011.110344

AR-11-0344

2

R Loffroy et al.

.................................................................................................................................................

This method is then called ‘prophylactic’, ‘blind’, or ‘empiric’ embolization. Although some authors have advocated the practice of endoscopy-directed embolization for angiographically negative acute hemorrhage from gastroduodenal ulcers (17 –19), hard evidence data from the literature to support its use are still lacking and this technique remains controversial. The purpose of our study was to retrospectively compare the results of arterial embolization for acute bleeding (BU) and non-bleeding (NBU) gastroduodenal ulcers.

Material and Methods Study cohort A retrospective chart review was conducted to identify patients admitted to our institution from January 2000 to February 2008 for transcatheter arterial embolization to treat massive bleeding (defined as a need for more than 4 U of blood/24 h) from gastroduodenal ulcers. These patients had re-bled despite initial conservative medical therapy and endoscopic treatment with local epinephrine injections or placement of metallic clips around the bleeding site. This retrospective study was performed in compliance with the requirements of our institutional review board. The need for written informed consent was waived. All patients were hospitalized secondary to upper gastrointestinal bleeding or experienced upper gastrointestinal bleeding from peptic ulcer at some point during their hospitalization. Most patients were at high operative risk because of associated co-morbid conditions. These included malignancy (n ¼ 21), coronary heart disease (n ¼ 19), pulmonary embolism or respiratory failure (n ¼ 18), hypertension (n ¼ 15), severe diabetes mellitus (n ¼ 14), heart failure (n ¼ 12), chronic renal failure (n ¼ 11), cardiac arrhythmia (n ¼ 10), stroke within the last two weeks (n ¼ 7), peripheral occlusive arterial disease (n ¼ 6), surgery within the last two weeks (n ¼ 4), and cirrhosis (n ¼ 3). Embolization technique All angiographic procedures were performed by 1 of 2 experienced interventional radiologists with standard percutaneous transfemoral catheterization using a 5-F or 6-F sheath. Selective opacification of the celiac trunk and superior mesenteric artery was performed routinely using a 5-F Simmons-type catheter (Cook, Bjaeverskov, Denmark), followed by superselective arteriography of the target vessel using a 2.9-F co-axial microcatheter (Progreat; Terumo, Leuven, Belgium). In 21 patients (BU group), active extravasation of contrast medium or a false aneurysm-like lesion was seen at the bleeding site. Among them, a vasospasm was also identified in 11 patients. Embolization of the target vessel was then performed as selectively as possible in all of these patients (Fig. 1). In the 36 patients with no identifiable bleeding site (NBU group), that is to say no extravasation or false aneurysm-like lesion, embolization of a particular target vessel was performed based on conclusive endoscopic examinations with identification of the source of bleeding (Fig. 2). In this group, a vasospasm was identified on

Fig. 1 Arterial embolization in an 87-year-old man with active bleeding (Forrest I) from a peptic ulcer in the fundus at endoscopy. (a, b) Extravasation of contrast medium from the left gastric artery at the celiac trunk and superselective angiography indicates active bleeding (arrows). (c) After arterial microcatheterization, bleeding was controlled after embolization of the left gastric artery using cyanoacrylate glue (arrows)

angiography in 8 patients. If endoscopy demonstrated a bleeding ulcer in the fundus or corpus, then the left gastric artery (LGA) was selectively embolized. If endoscopy

AR-11-0344 A comparison of the results of arterial embolization for bleeding and non-bleeding gastroduodenal ulcers

3

.................................................................................................................................................

were then selectively embolized on both sides of the bleeding site to prevent ‘back-door’ bleeding. When catheterization was possible, the right gastric artery was also embolized. Embolization was usually achieved using different embolic agents in combination or alone at the discretion of the interventional radiologist. Embolic agents used included a spectrum of metallic coils (0.035-inch steel coils or 0.018-inch soft platinum multiple-curled microcoils (Cook, Bjaeverskov, Denmark)), mechanically disrupted gelatin powder sheet (Curaspon; CuraMedical, Amsterdam, The Netherlands), cyanoacrylate surgical glue (Glubran; GEM SRL, Viareggio, Italy) mixed with ultrafluid lipiodol (Therapex; E-Z-EM, Montreal, Canada) in a 1:3 ratio, or calibrated microspheres (EmboGold; Biosphere Medical, Roissy Ch De Gaulle, France). A single embolic agent was used in 28 patients in our series. Two or three different embolic agents were used in 26 and two of the remaining patients, respectively. All agents were released near the bleeding site until cessation of extravasation of contrast and/or occlusion of the targeted vessel as shown by angiographic monitoring. Collected outcome data and patient follow-up A physical examination was performed during the hospital stay and 1 month after hospital discharge as part of routine patient care. Data on subsequent events were collected during telephone interviews of patients and their respective referral physicians. The following variables were collected from patients’ medical records: procedural success (defined as complete occlusion of the target vessel), 30-day clinical success (defined as the absence of rebleeding that required further intervention in the form of second embolization, re-endoscopic treatment, or surgery), need for 30-day and/or in-hospital re-intervention, number of total packed red blood cells transfused, length of hospital stay after embolization, complications (classified as major if they required surgery and/or prolonged hospitalization and as minor otherwise), short- and long-term mortality, cause of death, and follow-up. Rebleeding was defined as bleeding with a greater than 2.0 g/dL decrease in the hemoglobin level and/or a lack of effectiveness of conservative medical treatment. Statistical analysis

Fig. 2 Arterial embolization in a 73-year-old man with recent bleeding (Forrest II) from a postbulbar duodenal ulcer at endoscopy. (a, b) Selective angiography of the celiac trunk and superselective arteriogram of the gastroduodenal artery (GDA) before embolization: no evidence of contrast medium extravasation. (c) Result after coil embolization of the distal and proximal GDA (arrows) (with gelatin sponge placed in the arterial trunk), based on endoscopic findings. No re-bleeding was reported

showed a bleeding ulcer in the antrum or duodenum, the entire gastroduodenal artery (GDA), right gastroepiploic artery (RGEA), and/or pancreaticoduodenal arches (PDA)

Univariate analyses were performed, first to compare the NBU and BU groups, and second to investigate whether clinical failure was associated with blind embolization. Categorical variables were compared between groups using the Chi-square test; count and continuous variables were compared using the non-parametric Wilcoxon rank-sum test. Logistic regression analyses between clinical success and non-outcome variables were performed with and without the blind embolization variable and were then compared using a likelihood-ratio test to evaluate whether blind embolization affects the relationship. A stepwise multivariate logistic regression analysis was performed, with a P value of 0.10 for removal, to determine the variables for which blind embolization was independently associated with clinical failure.

AR-11-0344

4

R Loffroy et al.

.................................................................................................................................................

All statistical analyses were performed by using STATA software (StataCorp. LP 2009; Stata Statistical Software: Release 11.0; College Station, TX, USA). A P value 0.05 was considered statistically significant.

Results Patient demographics, endoscopic findings and angiographic characteristics are summarized in Table 1. Sex ratio was significantly different between patients who underwent embolization for NBU and those who underwent embolization for BU (P ¼ 0.047). There were a greater number of patients with prior coagulopathy in the NBU group although the P value indicated only marginal significance (P ¼ 0.059). The two groups did not differ for the other variables such as age, co-morbidities and antiinflammatory medication. There were no significant differences between the two groups for the following variables: number of endoscopies before embolization, bleeding source at endoscopy, and endoscopic Forrest’s classification of peptic ulcers (20). Regarding angiographic characteristics, mean time from bleeding onset to referral and types of vessels to be embolized did not differ between the two groups. Endovascular treatment was feasible in all of the 57 patients, a 100% technical success rate. Post-embolization outcomes are summarized in Table 2. Thirty-day clinical success was achieved in 13 of the 21 patients (61.9%) in the NBU group and 27 of the 36 (75%) in the BU group (P ¼ 0.297). Overall, in-hospital re-intervention rates for rebleeding were similar between the two groups

(P ¼ 0.419). The need for second embolization, re-endoscopic treatment, and subsequent surgical repair did not differ between the two groups (P ¼ 0.897, 0.095, 0.414, respectively). The total packed red blood cells, length of hospital stay and minor complication rate were also similar for the NBU and BU groups (P ¼ 0.219, 0.980, 0.572, respectively). The four minor complications reported were as follows: in a patient with proximal celiac trunk occlusion and retrograde filling of the GDA through the inferior pancreaticoduodenal artery, catheterization distal to the bleeding site was not feasible. In this case, retrograde occlusion of the inferior pancreaticoduodenal artery, GDA, and common hepatic artery was achieved using resorbable particles (after verification of portal venous flow). The liver enzyme activities increased in this patient and then returned to normal within a few days. Another patient had a transient increase in serum amylase levels without symptoms. In the last two patients, a microcoil migrated into a distal branch of the left hepatic artery during embolization; this event had no detectable angiographic, biological or clinical adverse effects. All major complications occurred in the empiric embolization group and consisted of two angiography-related complications. In one patient, a hematoma in the groin area requiring vascular surgery, and in the other patient, a false aneurysm of the femoral artery treated medically by compression. No ischemic gastrointestinal complications were identified by clinical examination. The two groups did not differ for short- and long-term mortality rates, cause of death, and length of follow-up. Furthermore, successfully treated patients were compared with the other patients. No significant association was

Table 1 Summary of patient demographics, endoscopic findings, and angiographic treatment characteristics Characteristic Patient demographics Age (years) Sex (M/F) Co-morbidities ¼ 1/.1 Prior coagulopathy AI medication Endoscopic findings Number of endoscopies Gastric/duodenal bleeding Forrest’s classification I: active bleeding II: recent bleeding III: no bleeding Angiographic characteristics Days to angiography Embolized vessel GDA (and/or PDA) only LGA only LGEA only GDA (and/or PDA and/or RGA) and RGEA

NBU group (n 5 21)

BU group (n 5 36)

P value

69.8 + 14.7 11 (52.4)/10 (47.6) 5 (23.8)/15 (71.4) 13 (61.9) 6 (28.6)

69.9 + 15.1 28 (77.8)/8 (22.2) 10 (27.8)/22 (61.1) 13 (36.1) 7 (19.4)

0.868 0.047 0.743 0.059 0.428

1.8 + 0.6 3 (14.3)/18 (85.7)

1.6 + 0.9 6 (16.7)/30 (83.3)

0.201 0.812 0.432

8 (38.1) 11 (52.4) 2 (9.5)

20 (55.6) 13 (36.1) 3 (8.3)

2.7 + 2.0

2.3 + 1.7

3 (14.3) 3 (14.3) 0 (0) 15 (71.4)

12 (33.3) 3 (8.3) 2 (5.6) 19 (52.8)

0.362 0.239

Data are mean + SD or n (%) P 0.05 was considered statistically significant International normalized ratio . 1.5 or partial thromboplastin time . 45 seconds, or platelet count , 80,000/mm3 AI ¼ anti-inflammatory: non-steroidal anti-inflammatory drugs or corticosteroids, NBU ¼ non-bleeding ulcer, NBU ¼ bleeding ulcer, M ¼ male, F ¼ female, GDA ¼ gastroduodenal artery, RGA ¼ right gastric artery, LGA ¼ left gastric artery, RGEA ¼ right gastroepiploic artery, LGEA ¼ left gastroepiploic artery, PDA ¼ pancreaticoduodenal arches


5

.................................................................................................................................................

Table 2 Comparison of outcomes of embolization between patients with and without bleeding ulcers Outcome 30-day clinical success Need for in-hospital re-intervention Second embolization Re-endoscopic treatment Need for subsequent surgery Total packed red blood cells Duration of hospitalization (days) Complications Minor Major Overall mortality Time Within 1 month After 1 month Cause Recurrent bleeding Underlying illness Follow-up (months)

NBU group (n 5 21)

BU group (n 5 36)

P value

13 (61.9) 8 (38.1)

27 (75) 10 (27.8)

0.297 0.419

1 (4.8)

2 (5.6)

0.897

6 (28.6)

4 (11.1)

0.095

1 (4.8)

4 (11.1)

0.414

9.5 + 6.7

11.7 + 7.9

0.219

16.9 + 15.3

18.7 + 22.1

0.980

2 (9.5) 2 (9.5)

2 (5.6) 0 (0)

0.572 0.059

6 (28.6) 4 (19.0)

9 (25) 7 (25.9)

0.768 0.574 0.286

2 (20)

1 (6.3)

8 (80) 16.6 + 20.3

15 (93.8) 23.2 + 27.7

0.413

Data are mean + SD or n (%) P 0.05 was considered statistically significant One patient needed in-hospital re-intervention after one month NBU ¼ non-bleeding ulcer, BU ¼ bleeding ulcer

found between clinical success and blind embolization, in either the univariate analysis or in the stepwise multivariate logistic regression analysis (odds ratio [OR] 0.54, 95% confidence interval [CI] 0.17– 1.72, P ¼ 0.30). Lastly, the likelihood-ratio test of our logistic regression analyses with and without blind embolization demonstrated that the relationship between clinical success and relevant non-

outcome variables was not affected by blind embolization, as seen in Table 3.

Discussion The use of endoscopy-guided arterial embolization for angiographically NBU is still debatable. Furthermore, most of the reported studies are limited in the number of patients and did not compare the main clinical outcomes of embolization for NBU with those of embolization for BU using statistical analysis (21). The main finding from the present study is that embolization can be performed successfully even when angiography fails to visualize extravasation of contrast medium. In our study, clinical outcomes were not different between patients who underwent embolization for NBU based on endoscopic findings and those who underwent embolization for BU after angiographic visualization of the active bleeding site. Furthermore, blind embolization, which was performed in 21 (36.8%) of the 57 patients, was not associated with the outcome of embolotherapy and did not affect the relationship between clinical success and non-outcome characteristics. Endoscopy, which is considered the gold standard to detect and treat a peptic ulcer, is reliable and diagnostic in the majority of cases, even during the acute bleeding episodes (2, 5–7). However, patients in whom hemostasis is not achieved need alternative therapy. Arterial embolization is now accepted as the salvage treatment of choice over surgery for acute bleeding from the upper gastrointestinal tract despite endoscopic treatment. Many published studies confirm the feasibility of this approach and the high technical and clinical success rates, ranging from 91–100% and from 63–100%, respectively, in all case-series over the last decade (8–11, 13, 14, 17–19). However, because upper peptic ulcer bleeding is intermittent in nature, extravasation of contrast medium may not be visualized on the arteriogram (17–19). In this situation, endoscopic findings are valuable to the interventional radiologist for determining the artery that requires occlusion. In these cases, the entire gastroduodenal or left

Table 3 Logistic regression analyses to evaluate the effect of blind embolization on relationships between clinical success and non-outcome characteristics Logistic Regression Analysis Without ‘blind embolization’

With ‘blind embolization’

Likelihood ratio test

Characteristic

OR

95% CI

P value

OR

95% CI

P value

P value

Age (years) Sex Co-morbidities Prior coagulopathy AI medication Endoscopies (n) Bleeding source Forrest’s classification Days to angiography Type of treated vessel

1.00 0.87 0.07 0.15 0.60 0.59 0.63 1.20 0.53 2.50

0.97 – 1.04 0.25 – 2.98 0.01 – 0.57 0.04 – 0.54 0.16 – 2.20 0.29 – 1.19 0.12 – 3.39 0.35 – 4.13 0.34 – 0.80 0.23 – 27.50

0.834 0.819 0.013 0.004 0.441 0.142 0.589 0.772 0.003 0.454

1.00 0.71 0.07 0.15 0.64 0.60 0.64 1.36 0.53 3.10

0.97– 1.04 0.19– 2.60 0.01– 0.59 0.04– 0.59 0.17– 2.38 0.30– 1.19 0.12– 3.50 0.38– 4.86 0.35– 0.81 0.26– 36.28

0.834 0.602 0.015 0.006 0.504 0.148 0.607 0.635 0.003 0.368

0.300 0.250 0.430 0.750 0.330 0.330 0.300 0.270 0.490 0.300

Likelihood ratio test of our logistic regression model with and without blind embolization was used to evaluate whether blind embolization affect the relationship between clinical success and each non-outcome characteristics P 0.05 was considered statistically significant International normalized ratio . 1.5 or partial thromboplastin time . 45 seconds, or platelet count , 80,000/mm3 OR ¼ odds ratio, CI ¼ confidence interval, AI ¼ anti-inflammatory: non-steroidal anti-inflammatory drugs or corticosteroids

AR-11-0344

6

R Loffroy et al.

.................................................................................................................................................

gastric artery is recommended to be embolized depending whether the bleeding site is endoscopically detected in the fundus, corpus, antrum, or duodenum (17–19, 21). Unlike lower gastrointestinal tract bleeding, angiographic confirmation of a bleeding site is not a prerequisite for transcatheter therapy in the upper gastrointestinal tract. Indeed, several succeeding studies showed no difference in outcomes between patients with negative and patients with positive angiography results treated with embolization (17, 22–26). Lang et al. (22) described 13 patients with massive upper gastrointestinal tract hemorrhage and normal angiogram results, and concluded that embolization of the left gastric artery in the absence of extravasation appears warranted when there is definite prior identification of a lesion in the left gastric artery territory. They further recommend it even if there is no prior localization of a lesion if the patient is at risk for multiple organ failure should bleeding recur. However, Lang’s study describes a relatively small sample size. In a series of 50 patients by Walsh et al. (23), there was no statistically significant difference in outcome between patients with positive angiograms and those with negative angiograms. In a recent larger series of 108 patients, Padia et al. (19) showed that arterial embolization was equally effective in patients who demonstrated active contrast medium extravasation at angiography as in those who did not show contrast extravasation. To the best of our knowledge, only three studies attempted to identify blind embolization as a factor affecting the outcome of embolization using multivariate logistic regression analysis. In the series from Schenker et al. (24), Aina et al. (17), and Poultsides et al. (25), embolization for NBU was performed in 103 (63%), 29 (38%), and 22 patients (39%), respectively, and was not found to independently predict angiographic failure. Unfortunately, in all these studies, the only variable that was evaluated was the clinical success, defined as the absence of significant rebleeding within 30 days. In our series, patients who were treated with embolization for NBU (21 of 57, 37%) did not have a statistically significant difference in the relevant outcomes compared with patients undergoing embolization for BU. Although prospective studies are needed to compare these two endovascular management strategies, our study suggests that embolization for NBU is as safe and effective as embolization for BU. There is a very low risk of ischemic complications after embolization in the upper gastrointestinal tract above the ligament of Treitz because the rich collateral supply to the stomach and duodenum. This allows for the systematic use of embolization in patients with angiographically negative acute hemorrhage from gastroduodenal ulcers. More controversial is the influence of the type of embolic agent on the clinical outcome. In our retrospective study, it was impossible to evaluate the selection of favorable embolic materials. Whatever the embolic agent used in the blind embolization group, no difference was found in the clinical outcomes between the blind and identifiable groups. We find that the use of cyanoacrylate glue is particularly of interest in angiographically positive hemodynamically unstable patients and in cases of underlying coagulopathy, because it provides faster and better hemostasis than other embolic agents. For non-bleeding ulcers, sandwich embolization was usually achieved with coils to

prevent ‘back-door’ bleeding in the GDA, and calibrated particles or gelatin sponge in the LGA. Embolization of the GDA or pancreaticoduodenal arch at the bleeding site is typically not difficult. However, we believe that in emergency cases with massive duodenal bleeding, complete and speedy hemostasis is more important to save lives than superselective embolization with calibrated particles. In total, we believe that further study is needed to recommend embolic materials, including gelatin sponge, particles, and glue, to be used in association with coils. Our study puts into question the need for other time consuming investigative methods considering the results from empiric embolization following endoscopic identification of the bleeding site. Since time is of essence, we feel multidetector-row computed tomography (MDCT) or pharmacoarteriography using intraarterial vasodilators, anticoagulants, and/or fibrinolytic agents, as advocated by some authors for localization of acute upper gastrointestinal tract bleeding (27, 28), may not be necessary. Indeed, although MDCT has a potential role in the evaluation of patients with acute upper gastrointestinal bleeding, our judgment is that any introduction of this procedure should be accompanied by careful prospective assessment. Moreover, MDCT may not be feasible to perform in hemodynamically unstable patients. Most of the controversy involves the need for nuclear scintigraphy before diagnostic angiography. The clinical presentation of gastrointestinal bleeding may be episodic or continual. The former ranges from minor episodes that resolve, to chronic intermittent bleeding, to severe lifethreatening hemorrhage. Angiography is warranted in the latter group and may, with provocation, play a diagnostic role in the second. Usually, nuclear scintigraphy is recommended before angiography in all cases of episodic bleeding. Angiography remains the primary diagnostic imaging tool in those patients with continual active hemorrhage, however, and delaying angiography for scintigraphy is not warranted. Wireless capsule endoscopy is now being used as a diagnostic tool for obscure gastrointestinal bleeding. At the present time, however, it is not a practical modality to use in acute upper gastrointestinal bleeding. Our study has some limitations. First, because of its retrospective nature, the technique of angiography and embolization was not standardized but at the discretion of the interventional radiologist. It may have influenced the angiographic and clinical outcomes. Furthermore, surgical and endoscopic interventions may vary widely in indication and technique, depending on the skills and experience of the involved colleagues. Control of these variables in a retrospective design is nearly impossible. Second, the favorable prognosis of blind embolized bleedings might be attributed to the fact that a certain proportion of patients with normal arteriographic findings had spontaneous bleeding arrest. However, because all patients in the study group had recurrent and/or endoscopic resistant hemorrhage, the likelihood of spontaneous cessation of bleeding was low in this population. Lastly, there is clearly a need for prospective, randomized, controlled studies comparing embolization of NBU and BU after endoscopic failure. In conclusion, our study showed no difference of outcome between patients who underwent embolization for NBU


7

.................................................................................................................................................

and those who underwent embolization for BU after failed attempts at endoscopic hemostasis. Therefore, in situations where no extravasation is seen angiographically, we advocate the practice of embolization of the most likely offending vessel based on endoscopic guidance. Conflict of interest: None. REFERENCES 1 Laine L, Peterson WL. Bleeding peptic ulcer. N Engl J Med 1994;331:717 – 27 2 National Institutes of Health. Consensus conference: therapeutic endoscopy and bleeding ulcers. JAMA 1989;262:1369 – 72 3 Rollhauser C, Fleischer DE. Nonvariceal upper gastrointestinal bleeding: an update. Endoscopy 1997;29:91 – 105 4 Ramsoekh D, van Leerdam ME, Rauws EA et al. Outcome of peptic ulcer bleeding, nonsteroidal anti-inflammatory drug use, and Helicobacter pylori infection. Clin Gastroenterol Hepatol 2005;3:859 – 64 5 Cheung FK, Lau JY. Management of massive peptic ulcer bleeding. Gastroenterol Clin North Am 2009;38:231 –43 6 ASGE Standards of Practice Committee, Banerjee S, Cash BD, Dominitz JA, et al. The role of endoscopy in the management of patients with peptic ulcer disease. Gastrointest Endosc 2010;71:663 –8 7 Enestvedt BK, Gralnek IM, Mattek N et al. Endoscopic therapy for peptic ulcer hemorrhage: practice variations in a multi-center U.S. consortium. Dig Dis Sci 2010;55:2568 – 76 8 Lang EK. Transcatheter embolization in management of hemorrhage from duodenal ulcer: long-term results and complications. Radiology 1992;182:703 – 7 9 Toyoda H, Nakano S, Takeda I et al. Transcatheter arterial embolization for massive bleeding from duodenal ulcers not controlled by endoscopic hemostasis. Endoscopy 1995;27:304 – 7 10 Loffroy R, Guiu B, Cercueil JP et al. Refractory bleeding from gastroduodenal ulcers: arterial embolization in high-operative-risk patients. J Clin Gastroenterol 2008;42:361 –7 11 Larssen L, Moger T, Bjornbeth BA et al. Transcatheter arterial embolization in the management of bleeding duodenal ulcers: a 5.5-year retrospective study of treatment and outcome. Scand J Gastroenterol 2008;43:217 –22 12 Loffroy R, Rao P, Ota S et al. Embolization of acute nonvariceal upper gastrointestinal hemorrhage resistant to endoscopic treatment: results and predictors of recurrent bleeding. Cardiovasc Intervent Radiol 2010;33:1088 – 100 13 Van Vugt R, Bosscha K, van Munster IP et al. Embolization as treatment of choice for bleeding peptic ulcers in high-risk patients. Dig Surg 2009;26:37 –42

14 Loffroy R, Guiu B, Mezzetta L et al. Short- and long-term results of transcatheter embolization for massive arterial hemorrhage from gastroduodenal ulcers not controlled by endoscopic hemostasis. Can J Gastroenterol 2009;23:115 – 20 15 Defreyne L, De Schrijver I, Decruyenaere J et al. Therapeutic decision-making in endoscopically unmanageable nonvariceal upper gastrointestinal hemorrhage. Cardiovasc Intervent Radiol 2008;31:897 –905 16 Miller M Jr, Smith TP. Angiographic diagnosis and endovascular management of nonvariceal gastrointestinal hemorrhage. Gastroenterol Clin North Am 2005;34:735 –52 17 Aina R, Oliva VL, Therasse E et al. Arterial embolotherapy for upper gastrointestinal hemorrhage: outcome assessment. J Vasc Interv Radiol 2001;12:195 –200 18 Loffroy R, Guiu B, D’Athis P et al. Arterial embolotherapy for endoscopically unmanageable acute gastroduodenal hemorrhage: predictors of early rebleeding. Clin Gastroenterol Hepatol 2009;7:515 –23 19 Padia SA, Geisinger MA, Newman JS, et al. Effectiveness of coil embolization in angiographically detectable versus non-detectable sources of upper gastrointestinal hemorrhage. J Vasc Interv Radiol 2009;20:461 –6 20 Forrest JA, Finlayson ND, Schearman DJ. Endoscopy in gastrointestinal bleeding. Lancet 1974;2:394 – 7 21 Holme JB, Nielsen DT, Funch-Jensen P, et al. Transcatheter arterial embolization in patients with bleeding duodenal ulcer: an alternative to surgery. Acta Radiol 2006;47:244 – 7 22 Lang EV, Picus D, Marx MV, et al. Massive upper gastrointestinal hemorrhage with normal findings on arteriography: value of prophylactic embolization of the left gastric artery. Am J Roentgenol 1992;158:547 – 9 23 Walsh RM, Anain P, Geisinger M, et al. Role of angiography and embolization for massive gastroduodenal hemorrhage. J Gastrointest Surg 1999;3:61 –5 24 Schenker MP, Duszak R Jr, Soulen MC, et al. Upper gastrointestinal hemorrhage and transcatheter embolotherapy: clinical and technical factors impacting success and survival. J Vasc Interv Radiol 2001;12:1263 – 71 25 Poultsides GA, Kim CJ, Orlando R 3rd, et al. Angiographic embolization for gastroduodenal hemorrhage: safety, efficacy, and predictors of outcome. Arch Surg 2008;143:457 – 61 26 Ichiro I, Shushi H, Akihiko I, et al. Empiric transcatheter arterial embolization for massive bleeding from duodenal ulcers: efficacy and complications. J Vasc Interv Radiol 2011;22:911 – 6 27 Bloomfeld RS, Smith TP, Schneider AM, et al. Provocative angiography in patients with gastrointestinal hemorrhage of obscure origin. Am J Gastroenterol 2000;95:2807 –12 28 Johnston C, Tuite D, Pritchard R, et al. Use of provocative angiography to localize site in recurrent gastrointestinal bleeding. Cardiovasc Intervent Radiol 2007;30:1042 – 6

QUERY FORM Royal Society of Medicine Journal Title: ACTA Article No: 11-0344 AUTHOR: The following queries have arisen during the editing of your manuscript. Please answer the queries by making the requisite corrections at the appropriate positions in the text. Query No.

Nature of Query NO Queries

Author’s Response