Abdominal EmilJ.

Balthazar,

J.

#{149} Alec

MD

Megibow,

MD

and #{149} Shari

Appendicitis : Prospective with High-Resolution CT’ Computed

tomography

(CT)

used

to prospectively patients with clinical

acute

appendicitis.

were

performed

was

evaluate indications

when

an abnormal

or inflammatory pendicolith

changes were

for

appendix

plus

detected.

an ap-

Failure

to

visualize an abnormal appendix or appendicolith in the presence of periceca! inflammatory changes was considered suspicious but nonspecific. CT results were correlated with surgical and pathologic results (74 patients) and other radiologic and clinical findings (26 patients). CT

helped

to diagnose

patients)

and

appendicitis

nonspecific

(64

right

lower

quadrant inflammation (five patients) and to rule out appendicitis (31 patients). CT had a 98% sensitivity, an 83% specificity, and a 93% accuracy. In 17 of 31 patients without CT evidence of appendicitis, other conditions explaining their symptoms

were

detected.

agnosis

is in doubt,

successfully acute

When

CT

can

di-

be used

patients

with

appendicitis.

Index terms: Appendicitis, dix, CT, 751.1211 Radiology

the clinical

to evaluate

1991;

751.291

Appen-

#{149}

180:21-24

From the Department of Radiology, 560 First Aye, New York, NY 10016. Received December 10, 1990; revision requested January 14, 1991; revision received February 1; accepted February 15. Address reprint requests to E.J.B. © RSNA, 1991 See also the in this issue.

editorial

by Brown

(pp

E. Siegel,

MD

#{149} Bernard

13-14)

attempt

MD

Evaluation

the diagnosis of acute appendicitis is usually made clinically, the symptoms and the laboratory values are often nonspecific and confusing. This leads to a high prevalence of unnecessary appendectomies variably reported in the surgical literature at 8%-30%, with an average of about 20% (1-3). In addition to the unnecessary laparotomy, longer hospitalization, and increased cost, the disease actually responsible for the patient’s symptoms remains unclear in two-thirds of patients after surgery (2,3). In an

Radiology

A. Birnbaum,

MATERIALS

LTHOUGH

100

Examinations with the terminal ileum and cecum filled with contrast material. Acute appendicitis was diagnosed

Gastrointestinal

to improve

diagnostic

This

AND

study

includes

tients suspected who underwent tions

in our

years.

There

female

METHODS

100

consecutive

of having abdominal

institution

in the

52 male

were

patients

who

pa-

appendicitis, CT examinapast

21/2

patients

ranged

and 48

in age

from

9

to 87 years (mean, 42 years). This was a selected group of patients who presented with lower abdominal pain and tenderness ing,

but without low-grade

either nausea and fever, or leukocytosis.

ical presentation and

radiographic

lion was requested the

judgment

confirma-

entirely

and

the consulting

on the basis of

clinical

experience

surgeon.

constituted

Clinatypical

was considered

or nonspecific,

vomit-

These

approximately

of

individuals

20%

of patients

accuracy, newer imaging modalities, including computed tomography (CT) (4-9) and sonography (10-14), have been used and reported. Recently several investigators have stressed the role of graded compression sonography in the diagnosis of acute appendicitis (10-12). These authors have recommended sonography as the imaging modality of choice in patients suspected of having acute appendicitis (15). The potential role of

operated on for acute appendicitis in the same period of time in our institution. CT

CT in the

discrepancies findings were

diagnosis

of appendicitis

and in the evaluation of patients who are thought to have acute appendicitis has been neglected. Since our original report (7), we have been impressed and encouraged by the assistance CT gives to diagnosing and evaluating acute appendicitis. We believe that familiarization with the expected radiologic findings, faster CT equipment, and refinements in our technique are the factors responsible for CT providing such assistance. These positive developments prompted us to perform a prospective evaluation of the accuracy, the sensitivity, and the role of high-resolution CT in the

evaluation

of patients

with

clinical suspicion of acute appendicitis. The results of our study are the subject of this article.

examinations

hours made

were

performed

within

48

of admission, and the diagnosis with these findings, based on the

original related

preoperative CT with the available

pathologic performed nations,

findings.

report, surgical

was corand

Appendectomy

in 74 patients, clinical follow-up

was

and other evaluations,

exami-

and diagnoses made at discharge were available in the remaining 26 patients. CT scans

were

retrospectively

particular

interest

reviewed,

in the

between present.

cases CT

with

in which

and

pathologic

Examinations ond

tems,

were performed on 2-secscanner; GE Medical SysMilwaukee) with sequential, rapid,

units

thin-section 700-800

(9800

scanning. of 1.2%

mL

meglumine tol-Myers

diatrizoate Squibb, New

orally

at least

bolus

intravenous

43%

Patients

received

diluted

1 hour

barium

before

iodinated

contrast followed

140 mL at 0.8 mljsec.

A

of 50 mL of

material by

BrisNJ)

scanning.

injection

at 1.5 mljsec,

or 2%

(Gastrografin; Brunswick,

an

was

given

additional

Conventional

10 x

10-mm or 10 x 12-mm collimation was used in the upper abdomen, followed by 5 x 8-mm sections in the lower abdomen and

loops trast

upper

pelvis.

and cecum material

and

When

were the

terminal

ileal

not filled with diagnosis

was

conin

doubt, a repeat examination was performed 15-30 minutes later slightly above and below the cecal caput with 5 x 5-mm collimation.

21

RESULTS The

citis

CT

was

either

diagnosis

based

of the

The

appendix

axis

eased

on

as a small

(4-20-mm fluid filled appendix

of

sets

of ab-

set is an abnor-

visualized

longitudinal

structure usually

two

first

appendi-

detection

following

normalities. mal

of acute

on the

cross

or

tubular

outer diameter), (Fig 1). The dishas

an

enhanced

slightly thickened wall ated with periappendiceal tory changes. Extensive,

and

and

is associinflammapericecal,

hazy, and streaky opacities ing phlegmons and poorly

1.

phlegmon, or retrocolic

lated and visualized.

(1) Typical

CT features

of acute

appendicitis.

Cross

section

of a distended

contains Findings

fluid are

and solid elements strongly suggestive

(arrows). of but

not

Appendix specific

or appendicoliths for acute appendicitis.

fluid-

(arrowlower quadencapsuwere not

or lo-

cation accompanied pendicolith. The

by a calcified apdetection of inflam-

matory

phlegmon,

changes,

1, 2.

filled appendix (arrow) shows enhanced wall and periappendiceal inflammation heads). Terminal ileum and cecum (C) are filled with contrast material. (2) Right rant abscess caused by ruptured appendix in a 63-year-old man. Abscess is poorly

representdefined

fluid collections constituting abscesses are seen in the more advanced cases. The second set of abnormalities involves a combination of localized in-

flammatory changes, abscess in a pericecal

2.

Figures

or abscess

adjacent

to cecum without visualization of an abnormal appendix or an appendicolith was considered sufficient to suspect but not diagnose ap-

pendicitis. On the basis of the original CT report, a diagnosis of appendicitis was made in 64 patients, and suggestive

but

nonspecific

right

lower

quadrant

inflammatory changes were reported in five cases (Table, Fig 2). Of the remaining 31 cases, 14 patients had nor-

mal CT results and 17 patients had findings that demonstrated a variety of pathologic

changes.

CT

examina-

tions enabled the following conditions to be detected: pelvic inflammatory disease (six cases), ileocolitis (four),

pyelonephritis

(two),

In patients

tis, the

confirmed findings positive

Surgical available

and

diverticu-

pathoin 74 cases,

60 cases of true-positive with CT, four cases of falseresults, one case of false-neg-

ative findings, and four negative results. Among patients with suggestive cific right lower quadrant tion (Fig 2), three patients

cases of truethe five but nonspeinflammahad appen-

CT findings. or perforated

There ap-

pendices and 10 normal appendices resected, for a surgical false-positive rate of 13.5%. Normal appendices were

resected

diagnosed 22

#{149} Radiology

in four

with

of the

64 cases

CT as appendicitis,

in

proved

was (94%),

appendici-

established and

with right

lower quadrant inflammatory changes were seen in 63 patients (98%). In 60 cases, the CT results were true-positive; a diseased appendix was detected in 45 (75%), appendicoliths in 17 (28%), and abscesses in 17(28%). Statistical analysis of the data shows a slight variation in the final results, depending on the significance attributed to the five patients with the suggestive but nonspecific inflammatory changes in the right lower abdomen

dicitis and two did not (Table). The decision to perform surgery in 74 patients was made mainly on the basis of clinical evaluation but was influenced by the were 64 inflamed

with

diagnosis

CT in 60 patients

litis (two), ureteral calculus (one), pancreatitis (one), and subhepatic abscess caused by a perforated duodenal ulcer (one). logic correlation,

two of the five cases initially demonstrating nonspecific right lower quadrant inflammatory changes, and in four of the 14 cases considered normal at CT scanning.

and

pelvis.

Thus

if these

cases

are omitted, CT yields 98% sensitivity, 88% specificity, and 95% accuracy in the diagnosis of acute appendicitis. If the suggestive inflammatory changes are considered diagnostic of appendicitis, CT yields 98% sensitivity, 83% specificity, and 93% accuracy. In both instances the positive and negative predictive values are well above 90% (94% and 97% in the first instance and 91% and 97% in the second).

DISCUSSION

In our institution the use of highresolution CT in patients suspected of having acute appendicitis has led to a substantial improvement in the preoperative diagnosis, as reflected by this prospective study. Compared with our previously reported study (7), the prevalence of CT visualization of a diseased appendix has increased from 18% to 75%, detection of appendicoliths from 23% to 28%, and the presence of typical findings indicative of appendicitis from 79% to 94%. These figures should be expected to vary in different series, depending on the type and severity of the pathologic process. The appendix is only slightly distended and thickened in milder forms (Figs 1, 3), whereas it is necrosed, ruptured, and replaced by a phlegmon or abscess in advanced forms of appendicitis (Figs 2, 4). In this series, abscesses were present in only 28% of patients, underlining the higher sensitivity of high-resolution CT

in the

diagnosis

of not

only

severe

forms but also mild incipient forms of appendicitis (Figs 1, 3). With an accuracy rate of 93% and a sensitivity of 98%, CT is far superior to barium enema examination and compares faJuly

i991

Figure 3. Acute appendicitis with penappendiceal phlegmon. Cross section of a distended fluid-filled appendix (arrowhead) with enhanced hyperemic wall. Note the extensive penappendiceal phlegmon (arrows) surrounding the appendix adjacent to the psoas muscle (P).

Figure 5. False-positive CT findings for appendicitis. Cecum (C) and terminal ileum are collapsed and not filled with contrast material. Adjacent loop of small bowel (arrow) is thought to represent inflamed appendix. Appendix was normal at surgery.

vorably with the reported data on graded compression sonography (1015). In addition to its high diagnostic accuracy, two other important factors emphasize the value of CT imaging in patients suspected of having acute appendicitis. First, CT can help in the accurate evaluation of the nature and extent of the pathologic process. It can differentiate mild inflammatory reaction from phlegmons and abscesses. It can reveal extension of disease far from the appendiceal region, within the pelvis and/or retrocolic and subhepatic space, and it can depict small amounts of extraluminal air associated with intraperitoneal or retroperitoneal perforations. Second, as evident in this study, CT can help in the diagnosis of intraabdominal disease unrelated to appendicitis, which explains the patient’s clinical symptoms. In our present series, 17 of the 31 patients suspected clinically of having appendicitis had other abdominal or pelvic abnormalities that were seen at CT examination. Volume

180

#{149} Number

1

a.

b.

Figure 4. Perforated appendix with pelvic abscess initially detected at sonography in a 39year-old woman. (a) Large fluid collection in pelvis is consistent with abscess (A). (b) Poorly encapsulated abscess in right side of pelvis (A) contains a calcified appendicolith (arrow). Appendix was not visualized. Surgery revealed ruptured and necrosed appendix with free-floating appendicolith in right lower quadrant abscess.

Analysis of the data in our prospective study reveals some of the remaining limitations and pitfalls of CT in the detection of appendicitis. Review of the CT scans showed that in two of the four cases of false-positive findings at CT, fluid-filled terminal ileal loops were misinterpreted as distended inflamed appendices (Fig 5). These errors could have been avoided if repeat examinations after proper filling of the terminal ileum had been performed. The importance of adequate visualization of the terminal ileum and cecum (filled with contrast material) cannot be overemphasized in the CT diagnosis of appendicitis. The third false-positive CT case showed a small amount of fluid adjacent to cecal caput, which was considered compatible with appendicitis (Fig 6). Review of the CT images revealed the presence of a normal appendix in the right lower quadrant, which was missed on the initial interpretation. A complex pericecal abscess was detected in the fourth false-positive case. Surgery revealed an adnexal abscess and a normal appendix. Finally, an erroneous interpretation was the cause of the only false-negative CT finding in this series. A distended fluid-filled appendix was misinterpreted as a fluid-filled ileal loop, despite adequate barium filling of adjacent ileal loops and cecum (Fig 7). Pericecal inflammatory changes or abscess collections in the retrocecal, retrocolic, right pelvic, or mesenteric locations without visualization of appendicoliths or a diseased appendix are further relative limitations of CT in the diagnosis of appendicitis (Fig 2). Although not specific, these find-

ings are suggestive of appendicitis and were present in five of our patients. In the absence of a convincing cause (history or other CT findings), a presumptive diagnosis of complicated appendicitis should be considered along with whether to perform a barium enema examination or surgery, depending on the clinical evaluation. In this series, among the five patients with suggestive CT findings, three had perforated appendicitis and two other patients had pelvic inflammatory disease. Analysis of our data reveals an 8.7% false-positive appendectomy rate if surgery had been performed only on the basis of the CT diagnosis of appendicitis and on the suggestive but nonspecific inflammatory right lower abdominal changes seen at CT. In this instance, however, one case of mild appendicitis would have been missed. Alternatively, if the suggestive findings had been overlooked and surgery performed only on patients with typical CT findings of appendicitis, the false-positive appendectomy rate would have been 6.2%, but four cases of appendicitis would have been missed. To avoid false-negative interpretations, a slightly higher rate of false-positive diagnoses should be accepted. In conclusion, CT, when properly performed and monitored, represents a valuable imaging modality in the detection and evaluation of acute appendicitis. Our results reveal a sensitivity of 98%, an accuracy of 93%, and a specificity of 83%, with predictive values of over 90%. In addition, CT can enable efficient evaluation of the nature of the pathologic process, reRadiology

#{149} 23

b.

a. Figure 6. False-positive (C) are not filled with

CT findings in a 24-year-old woman. (a) Terminal ileum and cecum contrast material. Small amount of fluid (arrow) is posterior to cecum to psoas muscle (P). CT findings were interpreted as compatible with appendiciinferior pelvic section revealed (retrospectively) a normal appendix (arrow). Sura ruptured right ovarian cyst and a normal appendix.

and

anterior tis. (b) More gery showed

veal extension of the disease far from the appendiceal region, and aid in the diagnosis of intraabdominal disease unrelated to appendicitis. The judicial use of high-resolution CT

will

diagnosis CT

can

lead

to a rapid

in most provide

and

cases. an

essential

road

1. 2.

3.

map

for the proper surgical or percutaneous abscess drainage and substantially reduce the number of appendectomies performed because of false-positive findings. At this time, we recommend the lution CT in patients

References

correct

Furthermore,

4.

6.

7. 8.

Acknowledgment: guerite A. McKnight tion.

The authors for manuscript

Berry JJr, Malt RA. Appendicitis near its centenary. Ann Surg 1984; 200:567-575. Bongard F, Landers DV, Lewis F. Differential diagnosis of appendicitis and pelvic inflammatory disease. AmJ Surg 1985; 150: 90-96. Lau WY, Fan ST, Ylu TF, et al. Negative findings at appendectomy. Am J Surg 1984; 148:375-378. Fish B, SmulewiczJJ, Barek L. Role of computed tomography in diagnosis of appendiceal disorders. NY State J Med 1981;

10.

11.

12.

13.

81:900-904. 5.

use of high-resofor whom the

diagnosis of acute appendicitis is in doubt and particularly in elderly mdividuals with suspicious but atypical clinical findings. #{149}

Figure 7. False-negative CT findings for acute appendicitis. Mild inflammatory changes (solid arrow) are present adjacent to barium-filled cecum (c). Fluid-filled loop of bowel anterior to right ureter (open arrow) was interpreted as terminal ileum. Appendix was inflamed and distended at surgery.

thank Marprepara-

Jones B, Fishman EK, Siegelman 55. Computed tomography and appendiceal abscess: special applicability in the elderly. Comput Assist Tomogr 1983; 7:434-438. Gale ME, Birnbaum 5, Gerzof SG, et al. CT appearance of appendicitis and its local complications. J Comput Assist Tomogr 1985; 9:34-37. Balthazar EJ, Megibow AJ, Hulnick D, et al. CT of appendicitis. AJR 1986; 147:705-710. Balthazar EJ, Megibow AJ, Gordon RB, et al. Computed tomography of the abnormal appendix. J Comput Assist Tomogr

14.

15.

Abu-Yousef MM, Bleicher JJ, Maher JW, et al. High-resolution sonography of acute appendicitis. AJR 1987; 149:53-58. PuylaertJBCM. Acute appendicitis: US evaluation using graded compression. Radiology 1986; 158:355-360. Jeffrey RBJr, Laing FC, Lewis FR. Acute appendicitis: high-resolution real-time US findings. Radiology 1987; 163:11-14. Adams DH, Fine C, Brooks DC. Highresolution real-time ultrasound: a new tool in the diagnosis of acute appendicitis. Am Surg 1988; 155:93-97. Jeffrey RB Jr. Laing FC, Townsend RR. Acute appendicitis: sonographic criteria based on 250 cases. Radiology 1988; 167: 327-329. Abu-Yousef Mlvi, Franken EA. An overview of graded compression sonography in the diagnosis of acute appendicitis. Semin Ultrasound CT MR 1989; 10:352-363.

1988; 12:595-601. 9.

Balthazar EJ, Gordon citis. Semin Ultrasound

BG. CT of appendiCT MR 1989; 10:

326-340.

24

#{149} Radiology

July 1991