Key words epiploic appendagitis - abdominal ultrasonography - CEUS - clinical awareness - clinical
acceptance
Introduction
Johann Wolfgang von Goethe (1749–1832) remarked in a conversation with the chancellor
of Saxe-Weimar-Eisenach, Friedrich v. Müller (1779–1849), “You only see what you already
know and understand” [1 ]. In medicine, this means that you cannot recognize, name, or treat conditions without
knowledge of disease phenomena. For genetically determined, chronic, life-threatening,
rare diseases, the term “orphan diseases” was introduced and, in this context, centers
for rare and undiagnosed diseases have been established [2 ]
[3 ]. In addition, rare, non-genetically caused, acquired diseases such as epiploic appendagitis
(EA), which are often not diagnosed or are misdiagnosed and therefore place an unnecessary
burden on the healthcare system, present a diagnostic challenge in clinical medicine
[4 ]. In an unpublished survey of internal assistants at a university hospital, 90% did
not know the clinical presentation of EA. In 902 patients investigated by computed
tomography (CT) because of abdominal pain, the frequency of EA was estimated at 1.3%,
with an incidence of 8.8 cases/million/year, according to a retrospective analysis
[5 ]. Moreover, in patients with primarily diagnosed diverticulitis, approximately 7%
of cases definitively had EA [6 ].
The epiploic appendages are fat pendants of the colon that originate from the serosal
surface [7 ]. An unfixed end of the appendices epiploicae and consequently increased mobility
can lead to spontaneous torsion with the development of infarction (EA) [8 ]
[9 ]. The clinical presentation of EA is strictly localized peritonitic pain, which is
more frequent on the left than on the right side and does not differ from the clinical
presentation of left-sided acute diverticulitis (AD) and right-sided acute appendicitis
(AA) [10 ]
[11 ]. For this reason, the initial referral diagnosis is incorrect in almost all patients
with a final diagnosis of EA [12 ]
[13 ]. These patients are usually afebrile and occasionally suffer from vomiting or diarrhea
[10 ]. In comparison with the differential diagnosis of AD, patients with EA are usually
younger [12 ]
[14 ]. Furthermore, patients with EA show no or only mildly increased inflammatory parameters,
such as a moderate increase in C-reactive protein (CRP), in comparison with patients
with AD and AA [11 ]. The symptoms of EA, which is a self-limiting disease, can persist for up to 7 days
in follow-up examinations [10 ]. Due to the non-specific clinical symptoms, the diagnosis can be made only by imaging
methods [9 ]
[11 ]
[15 ].
Computed tomography imaging is considered to be the highest priority imaging in surgery
to clarify acute abdominal pathologies [16 ]. It has already been shown that EA can be diagnosed by CT and differentiated from
AD [15 ]
[17 ]. However, the rarity of the disease combined with the low awareness of EA leads
to a high rate of CT misdiagnoses even by radiologists [4 ]. Rao et al. reported in a retrospective study that 64% of cases were overlooked
in CT examinations [4 ]. However, ultrasound should be used in non-critical patients for better efficiency
and targeted use of CT examination [16 ]. Even in rare congenital and acquired abdominal diseases, ultrasound patterns have
already been described [18 ]
[19 ]
[20 ]
[21 ]
[22 ]
[23 ]
[24 ]
[25 ]
[26 ].
In 2002, Hollerweger et al. described the characteristic features of EA in B-mode
ultrasound (B-US) and color Doppler sonography (CDS) [9 ]. In B-US, EA appears as an echogenic, non-compressible lesion in real-time examination
that is located adjacent to the colon and is adherent to the abdominal wall [9 ]
[21 ]. In CDS, the lesions show a lack of central color flow due to a perfusion deficit
caused by the infarction [9 ]. As another noninvasive method for the diagnosis of perfusion disorders, contrast-enhanced
ultrasound (CEUS) is routinely established in clinical practice [27 ]. Görg et al. demonstrated the absence of central enhancement of the echogenic lesions
on CEUS and proved this to be a typical pattern for EA on CEUS [11 ]. Based on this data, ultrasound seems to be a potential “gold standard” for the
diagnosis of EA [9 ]
[11 ]
[21 ]. In 2019, the EFSUMB Gastrointestinal Ultrasound (GIUS) Task Force Group published
guidelines for rare gastrointestinal diseases and described a high level of agreement
regarding the ultrasound characteristics of EA [21 ].
This study aims to describe the clinical awareness and acceptance of ultrasound-diagnosed
acute EA and their importance to avoid unnecessary therapeutic and imaging measures
over a period of 17 years in a university hospital.
Patients and Methods
The data of 54 consecutive primarily outpatients with acute, localized, peritonitic
pain and EA diagnosed by B-US and CEUS examination according to the EFSUMB guidelines
from November 2003 to September 2020 were retrospectively analyzed [27 ]. The 54 study patients had a mean age of 41.1 years (range: 10–80 years, standard
deviation: 17 years). 37 patients were male, and 17 patients were female The data
from 15 of these patients were published in a previous report from our group [11.
All examinations were performed by a German Society for Ultrasound in Medicine (DEGUM)
Level III qualified examiner. Diagnosis of EA was indicated on B-US by an echogenic,
non-compressible lesion in real-time examination adjacent to the colon, adhering to
the abdominal wall ([Figa. 1a ] and 1b) and on CEUS by marked contrast enhancement of the lesion with a central
area of non-enhancement ([Fig. 1c ] and 1d) [9 ]
[11 ].
Fig. 1 A 20-year-old male patient with acute left lower abdominal pain in the previous three
days and a normal CRP value (<5 mg/l). a B-mode US image shows an echogenic, non-compressible lesion adjacent to the air-filled
sigmoid colon. b Illustration of the manifestation of EA in the B-mode US image in image A. Arrowheads
indicate the echogenic, non-compressible lesion as a typical B-mode US pattern of
EA. The lesion is adjacent to the air-filled sigmoid colon (*). c CEUS after 33 s shows inhomogeneous enhancement with a small central area of non-enhancement,
surrounded by hyperenhancement of the inflammatory fatty tissue. d Illustration of the manifestation of EA on the CEUS image in image C. Arrowheads
indicate the hyperenhanced inflammatory fatty tissue. The centrally located infarcted
fat tissue shows non-enhancement (*). The air-filled sigmoid colon is bordered in
red.
Furthermore, a sonographic and clinical follow-up was performed in 45/54 (83.3%) cases.
In all of these cases, the diagnosis of EA was confirmed by documented regression
of clinical symptoms and sonographic pathology without the occurrence of alternative
diagnoses. The following clinical, laboratory, and imaging data of the patients were
evaluated retrospectively. Furthermore, the clinical awareness and acceptance of EA
diagnoses were determined in all study patients based on the documentation generated
by treating physicians. In 2013, an educational training program regarding the diagnosis
and treatment of EA was initiated. Therefore, the study period was divided into two
segments (before 2013 and from 2013 onwards), and the clinical and imaging data were
compared between subgroups diagnosed before 2013 (n= 22) and from 2013 onwards (n=32).
Clinical Data
Clinical awareness
Clinical awareness of EA was defined as considering EA as a possible differential
diagnosis and was evaluated based on documented suspected clinical diagnosis upon
admission of the patient to the hospital.
Clinical and laboratory manifestations
Initial basic ultrasound examination
Frequency of performance and documented diagnoses of an initial basic ultrasound examination.
Computed tomography and magnetic resonance imaging data
Frequency of performance, date of examination in comparison with reference ultrasound,
and documented diagnoses by a CT or magnetic resonance imaging (MRI) examination.
Clinical acceptance
Clinical acceptance of EA was defined as accepting sonographically diagnosed EA by
a DEGUM level III qualified examiner as the final diagnosis. The difference between
the finally documented ultrasound diagnosis of EA by a DEGUM level III qualified examiner
and the documented definitive clinical diagnosis by the treating physician at the
time of the patient’s discharge from the hospital was used to determine the clinical
acceptance of the final sonographic diagnosis.
Treatment data
Statistical analysis
Statistical analysis was carried out with Fisher’s exact test. A p -value of <0.05 was defined as significant.
Results
In all study patients, EA was sonographically detected and diagnosed by a DEGUM level
III qualified examiner.
Clinical awareness
[Table 1 ] presents the clinically suspected diagnosis at the time of admission of the patient
to the hospital for all study patients. In one case (after 2013), EA was stated as
a suspected clinical diagnosis (1.9%).
Table 1 Clinical awareness of EA at admission (54 patients with sonographically diagnosed
EA).
Clinically suspected diagnosis n =54 (100%)
All study patients n= 54 (100%)
Study patients before 2013 n =22 (40.7%)
Study patients from 2013 onwards n =32 (59.3%)
Epiploic appendagitis
1 (1.9%)
0 (0%)
1 (3.1%)
Diverticulitis
39 (74.1%)
16 (72.7%)
24 (75.0%)
Appendicitis
7 (13.0%)
4 (18.2%)
3 (9.4%)
Gastroenteritis
2 (3.7%)
1 (4.5%)
1 (3.1%)
Colitis
2 (3.7%)
1 (4.5%)
1 (3.1%)
Cholecystitis
1 (1.9%)
0 (0%)
1 (3.1%)
Adhesive ileus
1 (1.9%)
0 (0%)
1 (3.1%)
Clinical and laboratory manifestations
All patients had strictly localized peritonitic pain at the site of the sonographically
detected pathology. The pain in the study patients was localized in the left lower
abdomen in 37/54 (68.5%), in the right lower abdomen in 6/54 (11.1%), in the left
upper abdomen in 5/54 (9.3%), in the right upper abdomen in 3/54 (5.6%), and in the
middle of the abdomen in 3/54 (5.6%) patients. 21/54 (38.9%) patients had a normal
CRP value (<5 mg/l), and 31/54 (57.4%) had a minimally increased CRP value with values
between 5 and 30 mg/l. Two (3.7%) patients had levels above 30 mg/l.
Initial basic ultrasound examination
In 20/54 (37%) patients, an initial basic ultrasound examination was performed before
the final ultrasound examination by a DEGUM level III qualified examiner with a diagnosis
of EA. The frequency of initial basic ultrasound examination was 6/22 (27.3%) before
2013 and 14/32 (43.8%) from 2013 onwards ([Table 2 ]). The frequency of initial basic ultrasound diagnostics was not significantly different
before 2013 and from 2013 onwards (p=0.262, Fisher’s exact test). In 2/20 (10%) patients,
EA was detected in basic ultrasound, and in 18/20 (90%) EA was not detected. The following
findings were documented during the basic ultrasound examination: 5/18 (27.8%) cases
with no pathology; 4/18 (22.2%) cases of bowel wall thickening; 2/18 (11.1%) cases
of sigmoid diverticulitis; 2/18 (11.1%) cases of an echogenic formation; 1/18 (5.6%)
case of a hypoechoic formation; 1/18 (5.6%) case of segmental inflammation of the
colon; 1/18 (5.6%) case of free abdominal fluid; 1/18 (5.6%) case of cholecystitis;
and 1/18 (5.6%) case of a cockade sign in the colon; 1/18 (5.6%).
Table 2 Diagnostic data (54 patients with sonographically diagnosed EA).
Diagnostic measures
All patient n= 54 (100%)
Study patients before 2013 n =22 (40.7%)
Study patients from 2013 onwards n =32 (59.3%)
Diagnosis of EA detected or confirmed
Initial basic ultrasound examination
20 (37%)
6 (27.3%)
14 (43.8%)
2/20 (10%)
Computed tomography
15 (27.8%)
9 (40.9%)
6 (18.8%)
5/15 (33.3%)
Magnetic resonance imaging
1 (1.9%)
0 (0%)
1 (3.1%)
1/1 (100%)
Follow-up
45 (83.3%)
20 (90.9%)
25 (78.1%)
45/45 (100%)
Computed tomography and magnetic resonance imaging data
In all patients, clinical suspicion of sigmoid diverticulitis was documented as the
indication for CT and MRI examinations. In total, n= 15/54 (27.8%) patients had a CT examination, (n= 9/22 [40.9%] before 2013; n= 6/32 [18.8%] 2013 onwards) [Table 2 ]. The frequency of performed CT examinations was not significantly different before
2013 and from 2013 onwards (p= 0.225, Fisher’s exact test). Computed tomography was initiated before (n= 8/15, 53.3%) or after (n= 7/15, 46.7%) the documented final reference ultrasound examination. In CT examinations,
the following findings were reported: n= 5/15 (33.3%) cases of EA; n= 4/15 (26.7%) cases with no pathology; n= 2/15 (13.3%) cases of diverticulitis; n= 1/15 (6.7%) cases of gastroenteritis; n= 1/15 (6.7%) cases of lymphadenitis mesenterica; n= 1/15 (6.7%) cases of suspected colon perforation; and n= 1/15 (6.7%) cases of colitis. An MRI examination was performed in n= 1/32 (3.1%) patient after 2013, and EA was diagnosed ([Table 2 ]).
Clinical acceptance
In a total of n= 39/54 (72.2%) cases, EA was documented as the final clinical diagnosis, and, in n= 15/54 (27.8%) cases, diagnoses other than EA were documented as the final diagnosis
at the time of patient discharge despite confirmation of the EA diagnosis by a DEGUM
level III qualified examiner ([Table 3 ]). Of these patients, n =7/15 (46.7%) had a normal CRP value (<5 mg/l), and n =8/15 (53.3%) had a minimally increased CRP value of between 5 and 30 mg/l. Furthermore,
a clinical and sonographic follow-up was performed in n =14/15 (93.3%) patients, and the diagnosis of EA was confirmed with clinical and sonographic
regression with no evidence of an alternative diagnosis.
Table 3 Clinical acceptance of EA as final diagnosis (54 patients with sonographically diagnosed
EA).
Final clinical diagnosis n =54 (100%)
All study patients n= 54 (100%)*
Study patients before 2013 n =22 (40.7%)
Study patients from 2013 onwards n =32 (59.3%)
Epiploic appendagitis
39 (72.2%)
12 (54.5%)
27 (84.4%)
Unclear abdominal pain
10 (18.5%)
7 (31.8%)
3 (9.4%)
Diverticulitis
3 (5.5%)
1 (4.5%)
2 (6.3%)
Gastroenteritis
1 (1.9%)
1 (4.5%)
0 (0%)
Colitis
1 (1.9%)
1 (4.5%)
0 (0%)
*In all study patients, epiploic appendagitis was sonographically detected and diagnosed
by a DEGUM level III qualified examiner.
Before 2013 in n= 12/22 (54.5%) and from 2013 onwards in n= 27/32 (84.4%) patients, the final clinical diagnosis of EA was documented and accepted
by the treating physician at the time of patient discharge ([Fig. 2 ]). Clinical acceptance was significantly higher from 2013 onwards compared with before
2013 (p <0.05, Fisher’s exact test).
Fig. 2 Clinical acceptance of EA as final diagnosis (54 patients with sonographically diagnosed
EA).
Treatment data
In 28/54 (51.9%) cases, the patients received symptomatic therapy with analgesics.
In 26/54 (48.1%) patients, therapy was initiated, including hospitalization (inpatient
admission) (n =23/54 [42.6%]) and antibiotic therapy (15/54 [27.8%]). In 4/54 (7.4%) patients, antibiotic
therapy was recommended despite the acceptance of EA in the final clinical diagnosis.
Hospitalization was necessary in 14/22 (63.6%) patients before 2013 and in 9/32 (28.1%)
patients from 2013 onwards ([Table 4 ]). The frequency of hospitalization was significantly lower from 2013 onwards than
before 2013 (p <0.05, Fisher’s exact test). Furthermore, antibiotic therapy was administered to 7/22
(31.8%) patients before 2013 and to 9/32 (28.1%) patients from 2013 onwards ([Table 4 ]). The frequency of the administration of antibiotic therapy was not significantly
different before 2013 and from 2013 onwards (p= 0.772, Fisher’s exact test).
Table 4 Therapeutic data (54 patients with sonographically diagnosed EA).
Therapeutic measures
All study patients n =54 (100%)
Study patients before 2013 n =22 (40.7%)
Study patients from 2013 onwards n =32 (59.3%)
Hospitalization
23 (42.6%)
14 (63.6%)
9 (28.1%)
Antibiotic therapy
15 (27.8%)
7 (31.8%)
8 (25.0%)
Follow-up
In 45/54 (83.3%) patients, clinical and sonographic follow-up was performed (20/22
[90.9%] before 2013; 25/32 [78.1%] 2013 onwards). The frequency of performed clinical
and sonographic follow-up was not significantly different before 2013 and from 2013
onwards (p= 0.283, Fisher’s exact test).
Discussion
This retrospective study investigated the clinical awareness and acceptance of sonographically
diagnosed EA in 54 consecutive patients diagnosed by a DEGUM level III qualified examiner
during a period of 17 years in a university hospital. Our results revealed that the
level of awareness about the disease EA is extremely low. Only in n= 1/54 (1.9%) patient was EA considered as a suspected diagnosis at the time of hospital
admission. Therefore, attending physicians must be alerted to the potential differential
diagnosis of EA in patients with corresponding symptoms for the targeted use of diagnostic
procedures. In accordance with previous studies, the study patients were more frequently
male and middle-aged adults in comparison with patients with AD [28 ]. All patients reported acute localized abdominal pain, often localized in the left
lower abdomen, with missing or mildly increased inflammation parameters in contrast
to severe increased inflammatory parameters in patients with similar clinical symptoms
but with AD [12 ]. These characteristic features have been described previously in several studies
and should prompt the physician to consider the disease of EA [4 ]
[10 ]
[11 ]
[12 ]
[14 ]
[28 ].
As shown in previous research, high-quality ultrasound in combination with CEUS and
clinical and laboratory data enables the diagnosis of EA with high levels of agreement
regarding the ultrasound characteristics of EA [9 ]
[11 ]
[15 ]
[17 ]
[21 ]. In our analysis, 15 of 54 patients underwent an additional CT scan without specific
questioning for the diagnosis of EA, with a correct diagnosis in 5/15 (33.3%) patients.
Furthermore, EA was detected in the initial, routine, basic B-US examination in only
2/20 (10%) patients. The missing sonographic diagnosis by the primary attending physician
indicates the high interobserver variability, one of the major limitations of ultrasound
diagnostics in general [29 ]. However, a further important reason could be the dependence of the knowledge about
the disease on the clinical and sonographic experience of the ultrasound examiners.
In 15/54 (28.3%) study patients with EA confirmed by a highly qualified ultrasound
examiner, the diagnosis of EA was not accepted and documented as a final clinical
diagnosis. This may indicate a generally low level of trust in ultrasound diagnostics
or a lack of awareness of the diagnosis of EA by the responsible physician. The most
frequently mentioned misdiagnoses were unclear abdominal pain (18.5%) and diverticulitis
(5.5%). Moreover, the lack of knowledge of the disease led to unnecessary therapeutic
measures in 26/54 (48.1%) patients. These patients underwent antibiotic therapy, hospitalization,
or both. Interestingly, despite confirming the EA diagnosis, in four patients in our
study, the treating physicians ordered antibiotic therapy as the therapy of choice.
This reinforces again the lack of awareness of EA among physicians as a self-limiting
disease that requires only symptomatic analgesic therapy [4 ]
[12 ]
[15 ]. However, we have observed a significant increase in the acceptance of the diagnosis
of EA among treating physicians from 2013 onwards compared with before 2013 (p <0.05, Fisher’s exact test). This could be interpreted as an indication of increasing
knowledge and acceptance of the sonographic diagnosis of EA, apparently due to further
training in our hospital.
This examination was performed in a university hospital with an interdisciplinary
ultrasound center that has been established for over 30 years, performs 15 000 ultrasound
examinations per year, and employs experienced examiners. It must be assumed that
in primary care hospitals without an interdisciplinary ultrasound center and with
less qualified examiners and a lower quality of ultrasound machines, the level of
knowledge of EA’s clinical presentation and its clinical acceptance are even lower
and the frequency of misdiagnosis is even higher. Therefore, misdiagnoses of rare
diseases like EA may not be a local, but rather a general problem in the health care
system. In addition, general knowledge and skills in gastrointestinal ultrasound should
be considered and promoted [29 ].
Our study had some limitations. First, sonographic and clinical follow-up could be
performed in only n= 45/54 (83.3%) patients. Despite the evidence of the importance of B-US and CEUS in
the diagnosis of EA, the awareness of this disease in physicians in practice remains
the essential requirement for diagnosis [28 ]. This is based on the legitimate assumption that sonography is characterized by
a limited overview, high interobserver variability, high inter-device variability,
examiner-dependent documentation of findings, and patient-related limitations due
to obesity and meteorism [16 ]
[30 ]. In everyday clinical practice, this results in varying credibility and acceptance
of “subjective” sonographic findings and “objective” CT findings [16 ]. Furthermore, in the retrospective evaluation, it could not be determined whether
the investigator was blinded to the results of other examinations. However, this fact
did not matter, because this study only investigated clinical awareness/acceptance
of EA.
Conclusion
In our retrospective study, we showed that awareness and knowledge regarding the disease
of EA are extremely low. The mild or absent inflammatory parameter is an important
feature of this disease and could provide helpful information for diagnosis. In the
event of clinical suspicion of EA, a highly qualified sonographer should first perform
a sonographic examination. If there is clinical uncertainty, a clinical sonographic
follow-up should be conducted. Early detection of this disease can prevent unnecessary
therapeutic and imaging measures. Therefore, this disease should be highlighted in
ultrasound training for medical students and emergency medicine physicians. It must
be recognized that the lack of knowledge about the clinical presentation of EA, the
low diagnostic acceptance by clinicians, and the high degree of misdiagnosis in imaging
present a general problem in the healthcare system.
