Gastrointestinal Imaging Original Research

Flicek et al.
Diaphragm Disease of the Small Bowel

Gastrointestinal Imaging
Original Research
Downloaded from www.ajronline.org by 180.246.56.48 on 03/02/17 from IP address 180.246.56.48. Copyright ARRS. For personal use only; all rights reserved

Diaphragm Disease of the Small

Bowel: A Retrospective Review
of CT Findings
Kristina T. Flicek1 OBJECTIVE. The purpose of this article is to report the CT findings of pathologically
Amy K. Hara2 proven diaphragm disease in the small bowel.
Giovanni De Petris 3 MATERIALS AND METHODS. A retrospective review identified 12 patients with
Shabana F. Pasha4 pathologically proven small-bowel diaphragm disease who underwent CT within 6 months of
Anitha D. Yadav4 surgical resection. Two radiologists, who were unblinded to pathologic and clinical findings,
evaluated CT examinations for imaging findings of disease extent, appearance, and location.
C. Daniel Johnson2
Clinical history and postoperative follow-up were also performed.
Flicek KT, Hara AK, De Petris G, Pasha SF, Yadav RESULTS. The most common presenting symptoms were abdominal pain (7/12 [58%]) and
AD, Johnson CD anemia (5/12 [42%]). Long-term use of nonsteroidal antiinflammatory drugs was documented
in 58% (7/12) of patients. The most common location of small-bowel diaphragms was the ileum
(8/12 [67%]). The CT findings were abnormal in 92% (11/12) of patients. The most common CT
findings were small-bowel strictures (11/12 [92%]) and focal (median length, 1 cm) bowel wall
thickening (8/12 [67%]). Other less common CT findings included mucosal hyperenhancement
(6/12 [50%]), small-bowel dilatation (5/12 [42%]), and video capsule retention (6/9 [67%]).
Postoperative follow-up in 11 patients found recurrent symptoms in four patients.
CONCLUSION. Small-bowel diaphragm disease should be considered in patients with
a history of long-term use of nonsteroidal antiinflammatory drugs, chronic abdominal pain,
and anemia who present with CT findings of short, symmetric ileal strictures and focal bowel
wall thickening.

Keywords: diaphragm disease, mucosal diaphragm

S
disease, nonsteroidal-induced enteropathy mall-bowel diaphragm disease is
a relatively recent clinical entity,
DOI:10.2214/AJR.13.10732
with the first report on it pub-
Received February 11, 2013; accepted after revision lished in the pathologic literature
April 29, 2013. in 1988 [1]. The most common cause is be-
1
lieved to be long-term use of nonsteroidal an-
School of Medicine, University of Arizona, Tucson, AZ.
tiinflammatory drugs (NSAIDs), which inhibit
2
Department of Radiology, Mayo Clinic, 13400 E Shea cyclooxygenase-1 [26]. It is thought that cy-
Blvd, Scottsdale, AZ 85259. Address correspondence to clooxygenase-1 inhibition results in reduced
C. D. Johnson (johnson.daniel2@mayo.edu). microcirculatory blood flow, localized
3
ischemia, ulcers, strictures, and, finally, muco-
Division of Anatomic Pathology, Mayo Clinic,
Scottsdale, AZ.
sal diaphragms [1, 712]. In patients with
long-term use of NSAIDs, up to 70% may de-
4
Division of Gastroenterology, Mayo Clinic, Scottsdale, AZ. velop some degree of small-bowel enteropa-
thy [1315]. A recent capsule endoscopy
This article is available for credit.
study found diaphragm disease in 2% of 120
WEB patients taking long-term NSAIDs [16]. Other
This is a web exclusive article. conditions that can result in diaphragm disease
include potassium intake, celiac disease, eosin-
AJR 2014; 202:W140W145
ophilic gastroenteritis, and radiation injury [7].
0361803X/14/2022W140 Pathologically, diaphragms appear as a disk
of tissue protruding circumferentially into the
American Roentgen Ray Society intestinal lumen, reducing the lumen to a small
diameter (Fig. 1). Diaphragms can be single
but are more commonly multiple. They are
composed of mucosa and submucosa, with-
out involvement of muscularis propria or se-
rosa. The tip of the diaphragm is often ul-
cerated, creating a craterlike ulcer bed that
sometimes confers a club-shaped appearance
to the diaphragm [7].
The noninvasive diagnosis of diaphragm dis-
ease is challenging, particularly with imaging,
because the lesions may flatten and be difficult
to identify [7]. To our knowledge, the CT find-
ings of diaphragm disease have been reported in
only a single case report to date [17]. The pur-
pose of this retrospective review is to describe
the spectrum of CT findings in pathologically
proven diaphragm disease in the small bowel.
Materials and Methods
Our institutional review board approved this ret-
rospective study and determined that it was in accor-
dance with the privacy rule of the HIPAA and that
the criteria were met for waived patient authorization.

W140 AJR:202, February 2014


Downloaded from www.ajronline.org by 180.246.56.48 on 03/02/17 from IP address 180.246.56.48. Copyright ARRS. For personal use only; all rights reserved
A B
Fig. 1Microscopic and macroscopic images of diaphragms.
A, Typical diaphragm consists of enlarged plica that has at its tip craterlike ulcer (H and E, 100) with underlying submucosal fibromuscular scarring pointing toward
lumen.
B, Immunostain for smooth muscle actin (200) stained brown shows variability of muscular component in submucosa of diaphragm, with numerous muscle fibers. Tip of
diaphragm can assume club-shaped appearance.
C, Macroscopic appearance of ileal resection with diaphragms is shown. Diaphragms can be easily identified (arrows) or may be less conspicuous (asterisks), making
their enumeration difficult in individual cases.
Patient Population
A retrospective search of the pathologic and
surgical databases at our institution over the past
8 years for key diagnoses of NSAID enteropathy
or diaphragm disease identified 15 patients. Three
patients were excluded who did not undergo CT or
who underwent CT more than 6 months before sur-
gical resection. Twelve patients (eight women and
four men; mean age, 68.5 years; age range, 5083
years) composed the study group. All 12 patients
had small-bowel resections with a pathologic diag-
nosis of diaphragm disease and with CT within 6
months of the pathologic diagnosis.
Clinical History
Clinical symptoms such as abdominal pain, di-
arrhea, and weight loss were assessed by reviewing
the electronic medical record. In addition, any his-
tory of NSAID use, presence of anemia, or histo-
ry of obstruction was recorded. Clinical follow-up
was performed by review of the electronic medical
record to determine whether symptoms resolved
after surgical resection.
Reference Standard
The pathologic findings from surgical resection
(number of strictures and length of disease) served
as the reference standard (diagnostic criteria) for
all cases, and these findings were compared with
those found at CT and, if performed, at capsule en-
doscopy. The small-bowel location (jejunum or il-
eum) was determined by the operative reports.
CT Technique
All CT examinations were performed using a 16-
or 64-MDCT scanner. Nine of the 12 examinations
used neutral or low-density oral contrast material
(enterography, n= 7; angiography, n= 2), two used
AJR:202, February 2014 W141
Diaphragm Disease of the Small Bowel
positive oral contrast material, and one had no oral
contrast material. Contrast-enhanced CT examina-
tions (11/12) used approximately 150 mL of iodinat-
ed contrast material at 3 mL/s; scanning started at 70
seconds for routine CT or at 50 seconds for CT en-
terography, with a slice thickness of 35 mm and the
creation of both axial and coronal reformatted im-
ages. The time between imaging and the subsequent
pathologic diagnosis ranged from 0 to 115 days.
CT Image Analysis
The findings from the original radiologic inter-
pretation of each examination were recorded. Each
examination was then retrospectively reviewed by
two board-certified experienced (practicing lon-
ger than 10 years) gastrointestinal radiologists who
were not blinded to the pathologic results. During
the consensus review, the two radiologists record-
Fig. 2Mucosal diaphragm image from pill camera
examination. Lumen is markedly narrowed by
circumferential web or diaphragm.
C
ed the length and location of bowel wall thickening,
the presence of increased mucosal enhancement,
strictures, and any bowel dilatation. Abnormal bow-
el wall thickening was defined as thickness great-
er than 3 mm. Abnormal mucosal enhancement was
defined as hyperenhancement compared with the
rest of the bowel wall and adjacent segments. Stric-
tures were defined as luminal narrowing of more
than 50% of normally distended bowel. Small-bow-
el dilatation was defined as small-bowel loops with
a diameter of 3 cm or greater.
Capsule Endoscopy Correlation
Capsule endoscopy was performed in nine of
the 12 patients. Small-bowel capsule findings were
recorded, as well as whether the examination was
complete or incomplete (Fig. 2).
Results
Results are summarized in Table 1.
Clinical Presentation
The most common presenting complaint
was abdominal pain (7/12). Other symptoms
included anemia (n= 5), diarrhea (n= 2),
weight loss (n= 2), stool positive for Hemoc-
cult (Beckman Coulter) (n= 2), and vomit-
ing (n= 2). Seven of the 12 patients (58%)
had documented long-term NSAID use, al-
though the duration and type of NSAIDs
could not be determined even after careful
review of the medical record. One patient
who denied using NSAIDs had been treat-
ed with radiation therapy for uterine cancer
11 years before the development of her ab-
dominal pain. One patient had long-term use
of opioids, and another patient had long-term
use of hydrocodone.
 Flicek et al.

Reference Standard Findings From Surgical Pathology

Retained, stricture, diaphragm

The diagnosis of diaphragm disease was confirmed pathologically after
small-bowel resection in all 12 patients. Pathology reports of all patients
Capsule Endoscopy

Retained, ulcer, stricture

showed diaphragms, and eight of the 12 patients (67%) had ulceration re-
ported. Diaphragm disease was found in the ileum in eight patients (67%)

Retained, stricture
Ulcer, diaphragm

Superficial ulcer
and in the jejunum in four patients (33%). The number of strictures per pa-

Retained, ulcer
Retained, ulcer
Ulcer, stricture
Downloaded from www.ajronline.org by 180.246.56.48 on 03/02/17 from IP address 180.246.56.48. Copyright ARRS. For personal use only; all rights reserved

tient ranged from one to more than 20. Seven patients had strictures extend-
Not done
Not done

Not done
Retained

ing over a 30-cm or longer segment of small bowel.

CT Findings
The most common retrospective CT finding was strictures in 11 of 12 pa-
tients (92%) that ranged in length from 5 mm to 10 cm (median length, 1 cm;
Small-Bowel
Dilatation

mean length, 2.8 cm). The next most common finding in eight of 12 patients
Present
Present
Present

Present
Present
(67%) was bowel wall thickening ranging in depth from 5 to 8 mm. Half of
the examinations showed mild-to-moderate mucosal hyperenhancement (6/12
TABLE 1: Radiologic-Pathologic-Endoscopic Correlation in 12 Patients With Diaphragm Disease and Follow-Up

[50%]) and half (6/12 [50%]) had prominent mesenteric lymph nodes, whereas
five of the 12 (42%) had mild small-bowel dilatation up to 3.8 cm in diameter
(Figs. 36). One patient had a retained capsule in the small bowel, implying a
Hyperenhancement

stricture, but the CT was otherwise negative without wall thickening or dilata-
Mucosal

tion; three strictures were identified pathologically. Bowel wall symmetry was
Present
Present
Present
Present

Present
Present
Unblinded CT Findings

maintained at the stricture in all cases. The terminal ileum was never involved.
Three of the 12 patients (25%) had engorged-appearing mesenteric vasculature.
The initial blinded CT reports had abnormal small-bowel findings in
eight patients. The retrospective CT review, when unblinded to pathology,
resulted in the identification of three additional abnormal CT small-bowel
examinations. Findings not reported on the initial interpretation included
Bowel Wall
Thickening

strictures with bowel wall thickening (n= 3) and mucosal hyperenhance-

Present
Present
Present
Present

Present
Present
Present
Present

ment (n= 2). None of the initial reports mentioned diaphragm disease in
the differential diagnosis.
In five patients, more strictures were found pathologically than were identi-
fied at CT. In one patient, two strictures were suspected at CT, whereas 18 were
found by pathology. The large discrepancy may have been because the CT was
Strictures

done without IV or oral contrast material. Three patients had more strictures
No. of

> 20

13
12

12

suspected at CT than found at pathology. The largest difference (13 CT and

2
3

2
5
0

1
3
1

one pathology) occurred in a patient with a history of pelvic irradiation who

had a retained capsule at a stricture. At surgery, 20 strictures were palpated in
the small bowel; thus, the difference may have been due to nonstenotic fibrotic
Resected Bowel

regions or underreporting of the other smaller strictures at pathology. The other

Length (cm)

3040

cases all had differences in stricture numbers of four or less.

83
30

60
32

37
20

20
52

15
6

Capsule Endoscopy Findings

Pathologic Findings

Six of the nine patients who underwent capsule endoscopy had capsule re-
tention and all required surgical removal. Six studies found ulcerations, four
Present
Present
Present
Present

Present
Present
Present
Present
Ulcers

found small-bowel strictures, and two found a mucosal diaphragm (Fig. 2).

Clinical Follow-Up
Clinical follow-up information was available for 11 of 12 patients for peri-
Strictures

Multiple

ods from 1 month to more than 18 months (median, 7 months) after surgical
No. of

> 20
13

18

resection. Seven of 11 patients (64%) had no symptomatic recurrence. Two

2
7

8
3
1
0
3
1

patients (18%) had a recurrence of anemia. Two had a small-bowel obstruc-

tion, and both underwent surgery. One was found to have another small-bow-
el diaphragm just distal to the initial anastomosis and the other was found to
Location
Disease

Jejunum

Jejunum

Jejunum

Jejunum

have diverticulitis.
aUnenhanced study.
Ileum
Ileum
Ileum
Ileum

Ileum

Ileum
Ileum
Ileum

Discussion
Diaphragm disease has been well documented in the gastroenterology and
Patient
No.

pathology literature [15, 7, 911], but the imaging findings of this entity have
10

12
11
6a
3
4
5
1
2

7
8
9

not yet been well evaluated. This may be because diaphragm disease was

W142 AJR:202, February 2014


Downloaded from www.ajronline.org by 180.246.56.48 on 03/02/17 from IP address 180.246.56.48. Copyright ARRS. For personal use only; all rights reserved
Fig. 378-year-old woman with bowel wall thickening.
A and B, Note long segment mucosal hyperenhancement (arrows, A) and luminal narrowing (arrows, B). Contrast-enhanced CT enterography revealed 20 small-bowel lesions.
commonly missed on small-bowel follow-
through examinations, although it is possible
that the previous reports of localized nonspe-
cific small-bowel ulcerations on small-bow-
el follow-through examinations represent the
same disease [18]. Most patients in our study
had no further symptoms after surgical resec-
tion, which indicates that the correct diagno-
sis and treatment provided clinical benefit.
In our practice, like many others, CT
has become the primary small-bowel imag-
ing test, far outnumbering barium exami-
nations of the small bowel. For this reason,
we wanted to identify the most common CT
features of diaphragm disease rather than
barium findings. In our study, the most fre-
quent finding at CT was a stricture (Figs.
36), most commonly within the ileum or
less commonly in the jejunum. The stric-
tures were typically multiple, and most ap-
peared as focal areas of stenosis with asso-
AJR:202, February 2014 W143
Diaphragm Disease of the Small Bowel
A B
Fig. 475-year-old
woman with focal
small-bowel strictures.
Small-bowel strictures
(arrows) and proximal
dilatation indicate partial
obstruction identified by
contrast-enhanced CT
enterography.
ciated bowel wall thickening (Figs. 3 and 4).
Focal regions of mucosal hyperenhancement
(Fig. 5) could represent areas of ulceration,
because ulcerations were also confirmed by
pathologic findings. An interesting feature
of diaphragm disease is that there are often
small-bowel strictures at CT without signif-
icant small-bowel dilatation. In our study,
fewer than half of the patients had bowel dil-
atation over 3 cm in diameter, and the largest
diameter was only 3.8 cm.
With just 12 patients, our study was too
small to allow determination of an optimal CT
technique for detecting strictures. The findings
can be identified radiologically, as shown by
the retrospective detection of strictures iden-
tified by both readers in 11 of the 12 patients.
The presenting signs and symptoms of anemia
and abdominal pain, and a history of small-
bowel obstructions and long-term NSAID use,
were commonly encountered in this patient
group. Radiologists should be alert to these
clues of anemia and abdominal pain (with or
without evidence of small-bowel obstruction)
and should carefully examine the images for
evidence of strictures. Although the imag-
ing findings can mimic those of Crohn dis-
ease, several distinct features differentiate dia-
phragm disease: first, the clinical presentation
does not typically include diarrhea; second, it
never involves the terminal ileum; third, mu-
cosal hyperenhancement is variably encoun-
tered; fourth, there is a predominance of fo-
cal strictures rather than long segment disease;
and fifth, bowel wall symmetry is preserved.
Retention of the capsule endoscope was
highly likely in these patients, occurring in
six of nine patients (67%). This reflects the
marked luminal narrowing and fibrotic nature
of these diaphragms, which differs from the
edematous thickening that can occur in oth-
er small-bowel inflammatory diseases. CT
enterography in patients with suspected dia-
phragm disease may be helpful before video
capsule endoscopy to identify strictures and
avoid capsule retention and emergent op-
erative intervention for bowel obstruction.
Discontinuation of NSAIDs is the primary
treatment and is associated with improve-
ment in symptoms. Small-bowel resection is
most commonly performed for the treatment
of symptomatic diaphragm disease [11], al-
though endoscopic balloon dilatation of the
strictures using double balloon enteroscopy
has also been reported [19, 20].
Diaphragm disease of the small bowel is
likely to be currently undiagnosed at CT. Al-
though most of the CT examinations were ini-
tially reported as abnormal, none raised the
 Flicek et al.

more focal regions of luminal small-bowel

narrowing, bowel wall thickening, capsule
retention, or mucosal hyperenhancement.
In many cases, there will not be significant
small-bowel dilatation. This entity can be
differentiated from Crohn disease by clinical
Downloaded from www.ajronline.org by 180.246.56.48 on 03/02/17 from IP address 180.246.56.48. Copyright ARRS. For personal use only; all rights reserved

history and imaging findings. Radiologists

should consider diaphragm disease in the
differential of patients with a strong history
of long-term NSAID use, chronic abdominal
pain, and anemia with strictures and bowel
wall thickening at CT.

References
1. Lang J, Price AB, Levi AJ, Burke M, Gumpel JM,
A B Bjarnason I. Diaphragm disease: pathology of
Fig. 576-year-old woman with diaphragm disease. disease of the small intestine induced by non-ste-
A and B, Contrast-enhanced CT enterography in coronal (A) and axial (B) planes shows focal area of stenosis roidal anti-inflammatory drugs. J Clin Pathol
(arrows) within normal caliber ileum. 1988; 41:516526
2. Bjarnason I, Williams P, Smethurst P, Peters TJ,
Levi AJ. Effect of non-steroidal anti-inflammatory
drugs and prostaglandins on the permeability of the
human small intestine. Gut 1986; 27:12921297
3. Bjarnason I, Williams P, So A, et al. Intestinal
permeability and inflammation in rheumatoid ar-
thritis: effects of non-steroidal anti-inflammatory
drugs. Lancet 1984; 2(8413):11711174
4. Bjarnason I, Zanelli G, Smith T, et al. Nonsteroi-
dal antiinflammatory drug-induced intestinal in-
flammation in humans. Gastroenterology 1987;
93:480489
5. Bjarnason I, Zanelli G, Smith T, et al. The patho-
genesis and consequence of non steroidal anti-in-
flammatory drug induced small intestinal inflam-
mation in man. Scand J Rheumatol Suppl 1987;
A B 16(S64):5562
Fig. 669-year-old woman with diffusely dilated small-bowel loops. 6. Jenkins RT, Rooney PJ, Jones DB, Bienenstock J,
A, Diffusely dilated small-bowel loops due to stenosis (arrows) are causing high-grade mechanical small-bowel Goodacre RL. Increased intestinal permeability
obstruction.
in patients with rheumatoid arthritis: a side-effect
B, Small-bowel wall thickening (arrow) can also be seen on contrast-enhanced CT enterography.
of oral nonsteroidal anti-inflammatory drug ther-
apy? Br J Rheumatol 1987; 26:103107
possibility of diaphragm disease as the cause. complete historical documentation made it 7. De Petris G, Lopez JI. Histopathology of dia-
Only two small-bowel strictures were reported difficult to definitively determine the cause of phragm disease of the small intestine: a study of
prospectively; most reports instead described diaphragm disease, although at least half of 10 cases from a single institution. Am J Clin
bowel wall thickening or mild dilatation only. the patients had reported long-term NSAID Pathol 2008; 130:518525
Awareness of this condition will likely be use. A third limitation is the unblinded na- 8. Levi S, de Lacey G, Price AB, Gumpel MJ, Levi
helpful for improved recognition in the fu- ture of the retrospective review, which likely AJ, Bjarnason I. Diaphragm-like strictures of
ture. The finding of a single diaphragm should led to an overestimation of the ability of im- the small bowel in patients treated with non-ste-
prompt a search for others, because they are aging to detect the disease. For this article, roidal anti-inflammatory drugs. Br J Radiol 1990;
often multiple. Care should be taken, howev- however, the goal was not to determine the 63:186189
er, because peristaltic contractions may mimic performance of radiologists in detecting di- 9. McCune KH, Allen D, Cranley B. Small bowel
the finding of a small-bowel diaphragm. aphragm disease but rather to determine the diaphragm disease: strictures associated with
There are several limitations to this study. spectrum of radiologic findings that can be non-steroidal anti-inflammatory drugs. Ulster
First, our patient population was small. A encountered in this condition. A future blind- Med J 1992; 61:182184
larger patient population is warranted to veri- ed study after radiologist education could be 10. Shumaker DA, Bladen K, Katon RM. NSAID-in-
fy results and more accurately assess the sen- done to evaluate performance. duced small bowel diaphragms and strictures di-
sitivity and specificity of radiologic imaging In conclusion, diaphragm disease can be agnosed with intraoperative enteroscopy. Can J
in detecting diaphragm disease. Second, in- detected at CT by identification of one or Gastroenterol 2001; 15:619623

W144 AJR:202, February 2014

 Diaphragm Disease of the Small Bowel

11. Slesser AA, Wharton R, Smith GV, Buchanan GN.
Systematic review of small bowel diaphragm dis-
ease requiring surgery. Colorectal Dis 2012;
14:804813
12. Bjarnason I, Takeuchi K. Intestinal permeability
in the pathogenesis of NSAID-induced enteropa-
Downloaded from www.ajronline.org by 180.246.56.48 on 03/02/17 from IP address 180.246.56.48. Copyright ARRS. For personal use only; all rights reserved
due to nonsteroidal anti-inflammatory drugs: a sur-
gical perspective. Colorectal Dis 2011; 13:463466
15. Velchuru VR, Rimal D, Studley JG, Sturzaker
HG. Diaphragm disease of the ileum. Int J Surg
2006; 4:184186
16. Maiden L, Thjodleifsson B, Seigal A, et al. Long-
23:4044
18. Carlson HC. Localized nonspecific ulceration of
the small intestine. Radiol Clin North Am 1969;
7:97103
19. Hayashi Y, Yamamoto H, Taguchi H, et al. Non-
steroidal anti-inflammatory drug-induced small-

thy. J Gastroenterol 2009; 44(suppl 19):2329
13. Bjarnason I, Hayllar J, MacPherson AJ, Russell
AS. Side effects of nonsteroidal anti-inflammato-
ry drugs on the small and large intestine in hu-
mans. Gastroenterology 1993; 104:18321847
14. Pilgrim S, Velchuru V, Waters G, Tsiamis A, Lal
R. Diaphragm disease and small bowel enteropathy
term effects of nonsteroidal anti-inflammatory
drugs and cyclooxygenase-2 selective agents on
the small bowel: a cross-sectional capsule enter-
oscopy study. Clin Gastroenterol Hepatol 2007;
5:10401045
17. Zalev AH, Gardiner GW, Warren RE. NSAID in-
jury to the small intestine. Abdom Imaging 1998;
bowel lesions identified by double-balloon endos-
copy: endoscopic features of the lesions and
endoscopic treatments for diaphragm disease. J
Gastroenterol 2009; 44(suppl 19):5763
20. Mehdizadeh S, Lo SK. Treatment of small-bowel
diaphragm disease by using double-balloon enter-
oscopy. Gastrointest Endosc 2006; 64:10141017

F O R YO U R I N F O R M AT I O N
This article is available for CME and Self-Assessment (SA-CME) credit that satisfies Part II requirements for
maintenance of certification (MOC). To access the examination for this article, follow the prompts.

AJR:202, February 2014 W145