Keywords DBE–double-balloon enteroscopy - OGIB–obscure gastrointestinal bleeding - red blood
transfusion - therapeutic endoscopy
Introduction
Obscure gastrointestinal bleeding (OGIB) is defined as a bleeding in the gastrointestinal
(GI) tract without a source identified with repeated GI evaluations and often requires
multiple transfusions of packed red blood cells (pRBCs).[1 ].[2 ]
[3 ]
[4 ] As much as 75% of the total OGIB is due to small bowel lesions.[5 ]
[6 ]
[7 ] Double-balloon enteroscopy (DBE) not only has a diagnostic potential but, due to
its therapeutic potential, also has the ability to alter outcomes.[8 ]
[9 ] However, there is a paucity of data regarding the clinical impact of therapeutic
DBE. To evaluate the impact of therapeutic DBE on red blood cell (RBC) transfusion
requirement in patients with obscure GI bleeding, we conducted a retrospective study
at two tertiary care centers in Wisconsin. Both of our centers utilized DBE (Fujifilm
Medical Systems Inc. USA–Endoscopy, Wayne, New Jersey USA) as the primary method of
balloon-assisted deep small bowel enteroscopy.
Methods
Patients
Data, from a prospective cohort database at two tertiary care medical centers, was
retrospectively analyzed on patients who underwent DBE procedure for indication, endoscopic
findings, therapeutic intervention, immediate complications, and RBC transfusion units
over a 55-month consecutive time period. A total of 211 DBE procedures were performed
during this time period, with 163 (77%) performed for the indication of OGIB; 142
antegrade DBE procedures, and 21 retrograde DBE. A total of 69 DBE patient cases (42%)
had reliable blood transfusion records, of which 61/69 (87%) were antegrade DBE. This
is due to the fact that the two centers are referral centers for DBE, and transfusion
records were not available in all cases from outside referring institutions. They
all underwent diagnostic or therapeutic intervention with DBE. When available, we
also obtained small bowel capsule endoscopy (CE) findings on those patients. Institutional
Review Boards at both University of Wisconsin and Aurora St. Luke’s Medical Center
approved the study. Informed consent was taken for DBE. In certain cases, CE result
was obtained from outside hospital and only report was available, and in other cases,
patients were referred directly for DBE, given the high index of suspicion for active,
obscure overt small bowel bleeding.
DBE was performed either anterograde or retrograde or both, based on the indication
and results of CE, when available, to assist with approximation of location of small
bowel lesion, distance of insertion, and route of access. All DBEs were performed
under general anesthesia; for retrograde DBE, patients underwent bowel preparation
the night before the procedure. Diagnostic biopsies or therapeutic procedures were
performed as appropriate. The site of most distal advancement was tattooed. DBE was
performed by four different experienced interventional endoscopists (DVG, MEB, AS,
and NMG).
Results
A total of 69 patients were included in this study. The mean age ± standard deviation
(SD) was 63 ± 17. There were 36.2% (25) females and 63.80% (44) males. All patients
had DBE evaluation for obscure GI bleeding as the primary diagnosis with 23/60 (38.3%)
presenting with active overt GI bleeding; additional secondary indications included
iron deficiency anemia, abdominal pain, weight loss, and diarrhea ([Table 1 ]). A total of 61 cases were per oral antegrade DBE, with mean insertion time of 78.6
minutes to maximum depth of insertion to region of proximal-to-mid ileum (range 180–240
cm distal to ligament of Treitz), and overall average procedure duration (insertion
and withdrawal) of 95.2 minutes. The remaining eight cases were per colon retrograde
DBE, with mean insertion time of 62.6 minutes to the distal ileum, total procedure
duration of 74.6 minutes, and maximum point of insertion into distal ileum approximately
110 to 120 cm proximal to the ileocecal (IC) valve.
Table 1
Indications (primary +/- secondary) in DBE for OGIB
OGIB (overt & occult + secondary) indication
Total
Number
69
%age
100
Abbreviations: DBE, double-balloon enteroscopy; OGIB, obscure gastrointestinal bleeding.
Primary overt OGIB
23
33.30
Iron deficiency anemia
22
31.90
Abdominal pain
18
26.10
Weight loss
3
4.30
Other etiology
2
2.90
Diarrhea
1
1.40
All patients referred for DBE had prior negative upper oesophagogastroduodenscopy
(OGD) and colonoscopy, which did not demonstrate source of GI bleeding. Tagged RBC
scan was done in 8.7% of the patients all presenting with active overt OGIB. CE reports
were available only in 55% patients, with findings as described in [Table 2 ]. CE data was not available in all patients, because some patients either went directly
to DBE due to active bleeding or had CE done at other facility. DBE was done in all
patients, with findings described in [Table 3 ]. Of the 20 cases reported as normal, 15/20 had per oral DBE advanced to region of
proximal/mid ileum at maximum point of insertion with tattoo placed, three were retrograde
to distal ileum, and two had bidirectional balloon-assisted deep enteroscopy (i.e.,
retrograde DBE after antegrade DBE demonstrated normal findings). The 20 normal cases
had small bowel biopsies performed, which were also within normal limits.
Table 2
Capsule endoscopy findings
CE
Number
% age
Abbreviations: AVM, arteriovenous malformation; CE, capsule endoscopy; DBE, double-balloon
enteroscopy; SMT, submucosal tumor.
a CE data was not available in all patients because some patients either went directly
to DBE due to active bleeding or had CE done at other facility.
AVMs
18
26.10
Normal
8
11.60
Other
5
7.20
Polyp/SMT/mass
4
5.80
Stricture
2
2.90
Ulcer
1
1.40
Data unavailablea
31
44.90
Total
69
100.00
Table 3
OGIB pathology findings on DBE
Findings
Number
%age
Abbreviations: AVM, arteriovenous malformation; DBE, double-balloon enteroscopy; GIST,
gastrointestinal stromal tumor; IBD, inflammatory bowel disease; NSAID, nonsteroidal
anti-inflammatory drug.
AVMs
22
31.9
Normal
20
29.0
Other:
15
21.7
– Crohn’s IBD jejunum &/or proximal Ileum
3
– Crohn’s IBD distal ileum
4
– Celiac disease: ulcerative/erosive jejunal-ileitis
4
– Celiac disease/gluten enteropathy
4
SMT/polyp/mass
– Mid-distal jejunal GIST
– Distal jejunal leiomyoma/lymphoma/adenocarcinoma
5
3
3 (1 each)
7.2
Ulcer
4
5.8
– NSAID
3
– Jejunal Dieulafoy bleed
1
Ulcerated midileum ischemic/inflammatory stricture
2
3.0
Small bowel diverticular bleed
1
1.4
Total
69
100.0
Therapeutic intervention was done in all arteriovenous malformation (AVM) patients
except one where AVM was subendothelial. The intervention consisted of hemostasis
with argon plasma coagulation (APC) and/or hemoclips. In five patients, polypectomy
was performed. Blood transfusion rates were calculated 6 months before and after DBE
as described in [Table 4 ]. Wilcoxon signed rank test statistics was used to find the difference in the rate
of blood transfusion, which showed statistically significant decrease in rate of blood
transfusion with a p value of < 0.001. We further looked in subcategories of ulcer ([Figs. 1 ]
[2 ]), AVMs ([Fig. 3 ]) and submucosal tumor (SMT)/polyp/mass ([Figs. 4 ]
[5 ]) and found that rate of transfusion was significantly reduced in patients with AVMs,
but not in those with ulcer or SMT/polyp/mass, as described in [Table 5 ].
Table 4
Comparison of blood transfusion rates
Blood transfusion required
6 months pre-DBE
6 months post DBE
Z valuea
p value*
Abbreviations: DBE, double-balloon enteroscopy.
a Wilcoxon signed rank test.
*p < 0.05; significant.
**p < 0.001.
0
39 (56.5%)
57 (82.6%)
4.594
< 0.001**
1–5
14 (20.3%)
11 (15.9%)
6–10
11 (15.9%)
1 (1.4%)
> 10
5 (7.2%)
–
Total
69
69
Table 5
Comparison of blood transfusion required before and after 6 months in patients with
AVMs
Findings
Blood transfusion required
Before 6 months
After 6 months
Z valuea
p value*
Abbreviation: AVM, arteriovenous malformation.
a Wilcoxon signed rank test = NS: p > 0.05; not significant.
*p < 0.05; significant.
**
p < 0.001; highly significant.
AVMs
0
4 (18.2%)
15 (68.2%)
3.508
< 0.001**
1–5
9 (40.9%)
7 (31.8%)
6–10
7 (31.8%)
–
> 10
2 (9.1%)
–
Total
22
22
Fig. 1 66 years/o female with overt obscure gastrointestinal bleeding (OGIB) presenting
with hematochezia. Retrograde double-balloon enteroscopy (DBE) showing active bleeding
ulcers in distal ileum.
Fig. 2 46 years/o male with occult obscure gastrointestinal bleeding (OGIB). Double-balloon
enteroscopy (DBE), demonstrating ischemic ulcerated stricture in midileum.
Fig. 3 69 years/o male with overt obscure gastrointestinal bleeding (OGIB). Antegrade double-balloon
enteroscopy (DBE), demonstrating arteriovenous malformation (AVM) in distal jejunum,
which is treated with cautery.
Fig. 4 46 years/o male with obscure gastrointestinal bleeding (OGIB) and chronic iron deficiency
anemia. Antegrade double-balloon enteroscopy (DBE), demonstrating submucosal tumor
(SMT) in the distal jejunum.
Fig. 5 81 years/o male with iron deficiency anemia and chronic obscure gastrointestinal
bleeding (OGIB). Antegrade double-balloon enteroscopy (DBE) with distal jejunal mass,
which is consistent with biopsy-proven small bowel adenocarcinoma.
Of the 15 reported as “other” DBE pathology findings, 7 had endoscopic and histopathology
features of erosion and superficial mucosal ulceration consistent with small intestinal
inflammatory bowel disease (Crohn’s distal ileum in 4, and jejunal/proximal ileum
Crohn’s in 3), 4 had ulcerative/erosive jejunoileitis in celiac disease without histologic
features of small bowel lymphoma, and 4 had celiac disease with patchy endoscopic
pattern of scalloped mucosa and subtotal villous atrophy without erosions or ulceration.
There were no immediate or delayed complications of DBE reported in this study.
Discussion
In this study, we found statistically significant reduction in the requirement of
blood transfusion after DBE (81.8% vs. 31.8%), due to therapeutic impact of DBE. Blood
transfusion requirements decreased significantly in those patients with small bowel
AVMs and those that needed > 5 pRBC transfusions prior to the procedure. Our findings
confirm the therapeutic benefit found in other similar studies, whereby coagulation
or mechanical hemostasis of AVMs results in decreased rebleeding and transfusion requirements.[6 ]
[10 ]
[11 ]
[12 ]
[13 ]
Although our data confirm some of the outcomes reported in other studies, our data
is unique since it mimics the “real-world” experience where DBE was performed by multiple
endoscopists, and followed-up patients postprocedure for at least 6 months for assessment
of outcomes.[14 ]
[15 ]
[16 ]
In our current study, we found that rate of transfusion was significantly reduced
in patients with AVMs as opposed to ulcers and SMT/polyps. While CE studies were only
done in 55% of cases, the source of GI bleed traced to AVMS was likely picked up by
CE, and DBE was performed with therapeutic intent to perform APC hemostatic therapy.
As treatment of AVMS anywhere in GI tract is likely to respond to coagulation therapy
(i.e., APC), it is expected that successful hemostasis would prevent recurrence of
GI bleeding and future requirement of blood transfusion should reduce significantly
postintervention.[17 ] DBE appears to have a durable clinical impact on patient care as demonstrated by
the reduced RBC transfusion requirement of patients up to at least 6 months following
the procedure.
It is interesting to note that DBE was normal in 29% of cases when performed for evaluation
of OGIB. Normal DBE and negative therapeutic intervention was described in another
earlier study, which noted that DBE had no diagnostic yield in 44% of cases and no
therapeutic intervention performed in 55% of cases. That retrospective case-control
study of 55 DBE cases had higher negative diagnostic yield and lower therapeutic rate
compared with our study but similarly noted that AVMs accounted for the majority of
the DBE pathology findings. The authors suggested that factors such as fewer blood
transfusion requirements, absence of AVMs or ulcers and, possibly, endoscopist fatigue
maybe predictors of negative diagnostic and therapeutic yield of DBE.[18 ] The other factors such as extent of DBE insertion, average procedure duration, as
well as diverse array of small bowel pathological lesions detected on DBE are similar
to other published studies.[19 ]
[20 ]
[21 ]
The main limitation of our study is it is a retrospective analysis of data and has
a relatively small sample size. This is in part was due to the pRBC transfusion data
which was available only in 42% of patients undergoing DBE for OGIB. Moreover, small
bowel CE findings were also unavailable in 45% of cases.
Nevertheless, our study demonstrates that DBE is a safe procedure that can be employed
to evaluate OGIB. Additionally, DBE has a durable clinical impact on patient care
as demonstrated by reduced RBC transfusion requirements of patients up to 6 months
following the procedure. Additional prospective, multicenter studies should be designed
to further investigate this relationship and assess outcomes on need for hospitalizations
or disease-specific mortality.