Surgical
Downloaded from www.ajronline.org by 117.253.174.3 on 10/16/15 from IP address 117.253.174.3. Copyright ARRS. For personal use only; all rights reserved
MICHAEL
and
CRADE,’
Pathologic
KENNETH
Correlation of Cholecystosonography Cholecystography
J. W. TAYLOR,’
ARTHUR T. ROSENFIELD,’ PATRICK MINIHAN2
To define the accuracy of varying uftrasonic patterns in the diagnosis of gallstones, the records of 145 patients with ultrasound examination of the gallbladder prior to cholecystectomy were reviewed. Three abnormal scan categories were established: category 1-shadowing opacities that move with gravity within the gallbladder lumen; category 2-nonvisualization of the gallbladder lumen; and category 3- nonshadowing opacities within the gallbladder lumen. The accuracy of these criteria for gallstone diagnosis was found to be 100%, 96%, and 61%, respectively. Overall accuracy was 96% for gallbladder disease, with a 4% false negative rate. Oral cholecystography demonstrated an accuracy of 93% in this series. A preoperative ultrasound diagnosis of gallstones should probably be limited to category 1 and 2 appearances
CASPER
S. de GRAAFF,’
within a well defined gallbladder the gallbladder lumen, often
shadowing contracted dowing
AND
lumen; (2) nonvisualization with high level echoes
of
and
in the area of the gallbladder fossa suggesting a gallbladder, with or without gallstones; (3) nonshaopacities
physiological were
and
then
within
the
distention compared
to surgical
Results Category
gallbladder
lumen;
of the gallbladder.
and
These
(4)
normal
categories
results.
of Ultrasound
1
The classical appearances of gallstones, as opacities within the lumen of the gallbladder which shadow distally and move with gravity, were seen in 92 of the i45 patients (fig. 1 All 92 proved to have gallstones at surgery. One patient had an apparent intraluminal opacity that shadowed but did not move with gravity. Although our conclusion at the time was that these scan findings were related to gallstones, the patient proved to have a perforated duodenal ulcer with inflammation and fibrosis surrounding the gallbladder fossa. Since this case lacked documented opacity movement, it is not included in the category 1 series, but is described to illustrate the need to fulfill exactly these criteria.
only.
).
Earlier reports on the usefulness of ultrasound in the diagnosis of gallbladder disease [1-4] have recently been supported by further evaluation of its accuracy. Leopold et al. [5] claimed an accuracy of 9i% and Bartrum et al. [6] reported 93%. However, these studies compared the ultrasound findings to oral cholecystography without always having definite surgical proof of the diagnosis. We review the accuracy of ultrasonic diagnosis of 145 consecutive patients with surgically and pathologically proven gallbladder disease.
Subjects
and
Category
In 24 patients, the gallbladder lumen was not visualized on ultrasound despite fasting. High level echoes with evidence of shadowing were sometimes found within the gallbladder fossa (fig. 2), suggesting a contracted gallbladder, with or without gallstones. In 23 of the 24 patients surgery revealed a diseased gallbladder containing stones. in this category an accuracy of 96% was achieved. The single false positive case had three normal oral cholecystograms prior to our examination, and even in retrospect fulfilled category 2 criteria. An essentially normal, acalculous gallbladder was found at surgery. In addition, one patient not included in this series (since the gallbladder was not removed) was found to have agenesis of the gallbladder, an extremely rare condition with a reported incidence of 0.03% [7]. The scan made prior to surgery met criteria for Category 2, since the gallbladder lumen was not visualized.
Methods
The pathology records of the Yale-New Haven Hospital were sequentially reviewed, and 145 patients who underwent cholecystectomy after ultrasonic evaluation were selected for study. The medical records and the results of oral cholecystography were reviewed for additional information. Ultrasonic examination was carried out using a commercially available Picker 80L or Searle Pho-Sonic unit at a frequency of 2.5
or
3.5
emergency
MHz.
Although
basis,
preparation
some
for
patients
were
an elective
seen
on
examination
an
in-
volved a fast for at least 8 hr prior to examination, thus physiologically distending the gallbladder. Serial sector scans of the right upper quadrant were carried out in the transverse, oblique, and longitudinal planes. The patients were examined in the ization
supine, of the
decubitus,
and
erect
gallbladder
and
to determine
positions
to optimize whether
the
2
visualintralu-
minal opacities moved with gravity. We did not attempt to determine clinical accuracy; the interpretations of the ultrasound examinations were obtained from the officially dictated preoperative reports. On the basis of these descriptions, each scan was placed in one of four categories: (1) presence of shadowing opacities which moved with gravity
Category
3
A total of i8 patients had the gallbladder lumen.
within
Received November 29, 1977; accepted after revision March 27. 1978. ‘Department of Radiology, Yale University School of Medicine, 333 Cedar
Street,
New
Haven,
Connecticut
06510.
nonshadowing These were
Address
reprint
opacities variously re-
requests
to M.
Crade. Department Am J Ro.ntgenol © 1978 American
of Medicine,
Yale
University
131 :227-229, August Roentgen Ray Society
1978
School
of Medicine,
New Haven,
227
Connecticut
06510.
0361 -803X/78/08-0227
$00.00
Downloaded from www.ajronline.org by 117.253.174.3 on 10/16/15 from IP address 117.253.174.3. Copyright ARRS. For personal use only; all rights reserved
228
CRADE
ET
AL.
.‘4
%
-
..
-
ported
2, two patients. 0 = diaphragm,
as “sludge,”
In area of gallbladder L= liver, K = kidney.
“gravel,”
was
this
(fig.
reports
included
one
patient
3).
or”inspissated
fossa
bile”
there
In
in the
whose
i7 patients
gallbladder
was
perforated
at the time of surgery. gallstones nor significant Eleven
of the
one, the stone notably difficult acalculous
The course
Another patient had neither cholecystitis. patients in category 3 had gallstones. In was lodged within the cystic duct, a area
scan
accurately.
Six
au.,.
.
,
others
had
reflective
echoes
(arrows)
with
distal
shadowing
(s). No gallbladder
after an adequate number of scans, the gallbladder can be judged to be normal. Since normal gallbladders are only rarely confirmed by surgery, the great majority of normal studies will have no surgical or pathologic follow-up. Those that do are exceptions and usually result from misdiagnosis or in the course of
found
incidental
surgery.
Of the i45 patients reviewed, ii had normal proved to have normal gallbladders without inflammation, while five had diseased organs stones. The 4% false negative rate probably errors due to tomographic technique.
scans. Six stone or containing represents
4 most
common
normal.
appearance
The
evidence of internal must be performed search
to
cholecystitis.
Category
in
are highly
the gallbladder judged to be significantly inflamed at surgery, and was confirmed by pathologic examination. This
ultrasound
.
a
,..
Fig. 2.-Category lumen demonstrated.
.-
of
position
lumen
of
is
well
echoes. Multiple with the patient small
opacities;
the
gallbladder
Results
is of
A total
of 92 patients
visualized
without
tomographic in varying
scans positions
and
none
cholecystography
when
are
ultrasound
of this
of Oral
study
had
examinations.
was
not since
Cholecystography
both
oral
Although
to evaluate this
has
the been
cholecystography the
primary
accuracy well
goal of oral
documented
Downloaded from www.ajronline.org by 117.253.174.3 on 10/16/15 from IP address 117.253.174.3. Copyright ARRS. For personal use only; all rights reserved
CORRELATION
OF
CHOLECYSTOSONOGRAPHY
Fig. 3. -Category 3. Transverse scan with patient in decubitus demonstrating collection of opacities (arrow) within gallbladder (G). No evidence of shadowing. K = kidney. L = liver.
position lumen
[9-u],
it is of interest that stones were positively visualized by cholecystography in only 25 of the 84 patients (30%) who had this study and were subsequently proven to have stones. Of the remaining 59 patients, 53 had nonvisualized or poorly seen galibladders (63%), and six had false normal cholecystograms. Eight patients without gallstones had normal oral cholecystograms. Many patients (37%) did not have oral cholecystography. This underlines the well recognized limitations of that procedure [12, i3] for patients with severe vomiting, diarrhea, and jaundice, and in some patients with acute cholecystitis who must proceed directly to surgery on the results of clinical and ultrasound examinations alone. Many routine cases were also brought to surgery without cholecystography, reflecting the growing confidence in and acceptance of ultrasound as a reliable and economic diagnostic tool.
AND
all with gallstones. In our 24 patients with this finding (category 2), a diseased organ with gallstones was found at surgery in all but one case (96% correlation with true pathology) and was found in i8% of the abnormal cases. The ultrasonic pattern of nonshadowing opacities within the gallbladder lumen (category 3) is not specific for gallstones; only 6i% of the i8 cases brought to surgery had gallstones. However, in i7 of i8 cases, the gallbladder was inflamed at surgery. In one patient refusing surgery, and therefore not included in our report, serial scans disclosed a progression from normal to nonshadowing opacities to a contracted gallbladder. Almost certainly this patient had stones, and the phase of nonshadowing opacities represented an inflamed gallbladder on its way to stone formation. We therefore feel confident that nonshadowing opacities, though not diagnostic of gallstones, do suggest an inflamed and possibly symptomatic gallbladder. Of the i37 patients with proven gallbladder disease, i32 had abnormal ultrasound studies according to the above criteria (specificity of 96% for gallbladder disease). Of the 84 patients in this series with gallstones who underwent oral cholecystography, 93% had evidence of abnormality on ultrasound, either nonvisualization of the gallbladder lumen or visualization of a stone within the lumen. It is therefore evident that ultrasound is a useful and accurate imaging method that can be highly predictive when category i and 2 criteria are met.
REFERENCES 1.
Doust the
BD,
0:
Maklad
RR, Herman B-scan
68 : i 246-1
BB,
graphic
252,
Harris
11 :405-409,
J, Sarti Gastroen-
W: Ultrasonic a
Liver,
G,
comparison
gallbladder
and
radio-
Radiology
.
and
spleen.
Radiol
C/in
1975
Amberg
ultrasonic
ventional
MW, Castagna
1975
K, Brocker
13:543-55,
Leopold
scale
of
1974 EN:
NorthAm
examination 1974
ultrasonography.
cholecystography:
Carlsen
B-mode
1 1 1 :643-647,
of gallbladder
terology
5.
Ultrasonic
Radiology
J, Lindstrom
Value
3. Goldberg
4.
NF:
gallbladder.
2. Tabrisky
Discussion
The overall accuracy of ultrasound evaluation of gallstones has been previously reported, but without surgical confirmation in all cases and without evaluation of the accuracy of the individual categories of scanning patterns found. Our review of i45 surgically removed and pathologically studied gallbladders demonstrates that different categories of scanning images have different diagnostic importance. Shadowing opacities which move the gravity (category 1) were the most frequent findings in gallbladders containing stones, present in 92 of 137 abnormal cases (67%). Such findings are virtually pathognomic of gallstones. All 92 patients with this pattern had stones at surgery. The contracted gallbladder whose lumen is not seen on ultrasonic examination, even after fasting, has been previously reported. Leopold et al. [5] noted six cases,
229
CHOLECYSTOGRAPHY
J,
Gosink
BB,
cholecystography:
radiographic
Mittelstaedt
C:
a comparison
techniques.
Radiology
Gray-
with
con-
121 : 445-448,
1976 6.
Bartrum
R,
graphic
Crow
HC,
Foote
cholecystography.
SR: N
Ultrasonic
Eng/
J
and
Med
radio-
296:538-541,
1977 7. 8.
Monroe
SE,
der.
Med
Cal
Mujahed
bladder 9.
Baker
A, HL:
normal
1956
Evans
Further
JA,
Whalen
studies
on the
74:239-245,
SF: Performance
South
Mujahed
Congenital
Radiology
10. Ochsner 11.
FJ:
85:422-423,
absence JP:
The
on oral cholecystography.
tography. phy.
Ragen
Z:
Factors
gallbladder.
12.
Fischer
13.
cholangiography. Zboralske FE,
HW:
cholecystographic
interfering
Physiologic
accuracy
and
of oral
1974
cholecys-
of cholecystogra-
1970 Radiol
the
opacification 1:183-185,
pharmacologic
Radiol C/in North Am 4:623-625, Ainberg JR: Cholecysto-cholestasis: error.
gall-
112:1-3,
1960
with
Gastrointest
gallblad-
non-opacified
Radiology
and reliability
J 63:1268-1272,
Med
of the
Am J Dig Dis 7:339-346,
of
a
1976 aspects
of
1966 cause 1962
of
Downloaded from www.ajronline.org by 117.253.174.3 on 10/16/15 from IP address 117.253.174.3. Copyright ARRS. For personal use only; all rights reserved
MICHAEL
and
CRADE,’
Pathologic
KENNETH
Correlation of Cholecystosonography Cholecystography
J. W. TAYLOR,’
ARTHUR T. ROSENFIELD,’ PATRICK MINIHAN2
To define the accuracy of varying uftrasonic patterns in the diagnosis of gallstones, the records of 145 patients with ultrasound examination of the gallbladder prior to cholecystectomy were reviewed. Three abnormal scan categories were established: category 1-shadowing opacities that move with gravity within the gallbladder lumen; category 2-nonvisualization of the gallbladder lumen; and category 3- nonshadowing opacities within the gallbladder lumen. The accuracy of these criteria for gallstone diagnosis was found to be 100%, 96%, and 61%, respectively. Overall accuracy was 96% for gallbladder disease, with a 4% false negative rate. Oral cholecystography demonstrated an accuracy of 93% in this series. A preoperative ultrasound diagnosis of gallstones should probably be limited to category 1 and 2 appearances
CASPER
S. de GRAAFF,’
within a well defined gallbladder the gallbladder lumen, often
shadowing contracted dowing
AND
lumen; (2) nonvisualization with high level echoes
of
and
in the area of the gallbladder fossa suggesting a gallbladder, with or without gallstones; (3) nonshaopacities
physiological were
and
then
within
the
distention compared
to surgical
Results Category
gallbladder
lumen;
of the gallbladder.
and
These
(4)
normal
categories
results.
of Ultrasound
1
The classical appearances of gallstones, as opacities within the lumen of the gallbladder which shadow distally and move with gravity, were seen in 92 of the i45 patients (fig. 1 All 92 proved to have gallstones at surgery. One patient had an apparent intraluminal opacity that shadowed but did not move with gravity. Although our conclusion at the time was that these scan findings were related to gallstones, the patient proved to have a perforated duodenal ulcer with inflammation and fibrosis surrounding the gallbladder fossa. Since this case lacked documented opacity movement, it is not included in the category 1 series, but is described to illustrate the need to fulfill exactly these criteria.
only.
).
Earlier reports on the usefulness of ultrasound in the diagnosis of gallbladder disease [1-4] have recently been supported by further evaluation of its accuracy. Leopold et al. [5] claimed an accuracy of 9i% and Bartrum et al. [6] reported 93%. However, these studies compared the ultrasound findings to oral cholecystography without always having definite surgical proof of the diagnosis. We review the accuracy of ultrasonic diagnosis of 145 consecutive patients with surgically and pathologically proven gallbladder disease.
Subjects
and
Category
In 24 patients, the gallbladder lumen was not visualized on ultrasound despite fasting. High level echoes with evidence of shadowing were sometimes found within the gallbladder fossa (fig. 2), suggesting a contracted gallbladder, with or without gallstones. In 23 of the 24 patients surgery revealed a diseased gallbladder containing stones. in this category an accuracy of 96% was achieved. The single false positive case had three normal oral cholecystograms prior to our examination, and even in retrospect fulfilled category 2 criteria. An essentially normal, acalculous gallbladder was found at surgery. In addition, one patient not included in this series (since the gallbladder was not removed) was found to have agenesis of the gallbladder, an extremely rare condition with a reported incidence of 0.03% [7]. The scan made prior to surgery met criteria for Category 2, since the gallbladder lumen was not visualized.
Methods
The pathology records of the Yale-New Haven Hospital were sequentially reviewed, and 145 patients who underwent cholecystectomy after ultrasonic evaluation were selected for study. The medical records and the results of oral cholecystography were reviewed for additional information. Ultrasonic examination was carried out using a commercially available Picker 80L or Searle Pho-Sonic unit at a frequency of 2.5
or
3.5
emergency
MHz.
Although
basis,
preparation
some
for
patients
were
an elective
seen
on
examination
an
in-
volved a fast for at least 8 hr prior to examination, thus physiologically distending the gallbladder. Serial sector scans of the right upper quadrant were carried out in the transverse, oblique, and longitudinal planes. The patients were examined in the ization
supine, of the
decubitus,
and
erect
gallbladder
and
to determine
positions
to optimize whether
the
2
visualintralu-
minal opacities moved with gravity. We did not attempt to determine clinical accuracy; the interpretations of the ultrasound examinations were obtained from the officially dictated preoperative reports. On the basis of these descriptions, each scan was placed in one of four categories: (1) presence of shadowing opacities which moved with gravity
Category
3
A total of i8 patients had the gallbladder lumen.
within
Received November 29, 1977; accepted after revision March 27. 1978. ‘Department of Radiology, Yale University School of Medicine, 333 Cedar
Street,
New
Haven,
Connecticut
06510.
nonshadowing These were
Address
reprint
opacities variously re-
requests
to M.
Crade. Department Am J Ro.ntgenol © 1978 American
of Medicine,
Yale
University
131 :227-229, August Roentgen Ray Society
1978
School
of Medicine,
New Haven,
227
Connecticut
06510.
0361 -803X/78/08-0227
$00.00
Downloaded from www.ajronline.org by 117.253.174.3 on 10/16/15 from IP address 117.253.174.3. Copyright ARRS. For personal use only; all rights reserved
228
CRADE
ET
AL.
.‘4
%
-
..
-
ported
2, two patients. 0 = diaphragm,
as “sludge,”
In area of gallbladder L= liver, K = kidney.
“gravel,”
was
this
(fig.
reports
included
one
patient
3).
or”inspissated
fossa
bile”
there
In
in the
whose
i7 patients
gallbladder
was
perforated
at the time of surgery. gallstones nor significant Eleven
of the
one, the stone notably difficult acalculous
The course
Another patient had neither cholecystitis. patients in category 3 had gallstones. In was lodged within the cystic duct, a area
scan
accurately.
Six
au.,.
.
,
others
had
reflective
echoes
(arrows)
with
distal
shadowing
(s). No gallbladder
after an adequate number of scans, the gallbladder can be judged to be normal. Since normal gallbladders are only rarely confirmed by surgery, the great majority of normal studies will have no surgical or pathologic follow-up. Those that do are exceptions and usually result from misdiagnosis or in the course of
found
incidental
surgery.
Of the i45 patients reviewed, ii had normal proved to have normal gallbladders without inflammation, while five had diseased organs stones. The 4% false negative rate probably errors due to tomographic technique.
scans. Six stone or containing represents
4 most
common
normal.
appearance
The
evidence of internal must be performed search
to
cholecystitis.
Category
in
are highly
the gallbladder judged to be significantly inflamed at surgery, and was confirmed by pathologic examination. This
ultrasound
.
a
,..
Fig. 2.-Category lumen demonstrated.
.-
of
position
lumen
of
is
well
echoes. Multiple with the patient small
opacities;
the
gallbladder
Results
is of
A total
of 92 patients
visualized
without
tomographic in varying
scans positions
and
none
cholecystography
when
are
ultrasound
of this
of Oral
study
had
examinations.
was
not since
Cholecystography
both
oral
Although
to evaluate this
has
the been
cholecystography the
primary
accuracy well
goal of oral
documented
Downloaded from www.ajronline.org by 117.253.174.3 on 10/16/15 from IP address 117.253.174.3. Copyright ARRS. For personal use only; all rights reserved
CORRELATION
OF
CHOLECYSTOSONOGRAPHY
Fig. 3. -Category 3. Transverse scan with patient in decubitus demonstrating collection of opacities (arrow) within gallbladder (G). No evidence of shadowing. K = kidney. L = liver.
position lumen
[9-u],
it is of interest that stones were positively visualized by cholecystography in only 25 of the 84 patients (30%) who had this study and were subsequently proven to have stones. Of the remaining 59 patients, 53 had nonvisualized or poorly seen galibladders (63%), and six had false normal cholecystograms. Eight patients without gallstones had normal oral cholecystograms. Many patients (37%) did not have oral cholecystography. This underlines the well recognized limitations of that procedure [12, i3] for patients with severe vomiting, diarrhea, and jaundice, and in some patients with acute cholecystitis who must proceed directly to surgery on the results of clinical and ultrasound examinations alone. Many routine cases were also brought to surgery without cholecystography, reflecting the growing confidence in and acceptance of ultrasound as a reliable and economic diagnostic tool.
AND
all with gallstones. In our 24 patients with this finding (category 2), a diseased organ with gallstones was found at surgery in all but one case (96% correlation with true pathology) and was found in i8% of the abnormal cases. The ultrasonic pattern of nonshadowing opacities within the gallbladder lumen (category 3) is not specific for gallstones; only 6i% of the i8 cases brought to surgery had gallstones. However, in i7 of i8 cases, the gallbladder was inflamed at surgery. In one patient refusing surgery, and therefore not included in our report, serial scans disclosed a progression from normal to nonshadowing opacities to a contracted gallbladder. Almost certainly this patient had stones, and the phase of nonshadowing opacities represented an inflamed gallbladder on its way to stone formation. We therefore feel confident that nonshadowing opacities, though not diagnostic of gallstones, do suggest an inflamed and possibly symptomatic gallbladder. Of the i37 patients with proven gallbladder disease, i32 had abnormal ultrasound studies according to the above criteria (specificity of 96% for gallbladder disease). Of the 84 patients in this series with gallstones who underwent oral cholecystography, 93% had evidence of abnormality on ultrasound, either nonvisualization of the gallbladder lumen or visualization of a stone within the lumen. It is therefore evident that ultrasound is a useful and accurate imaging method that can be highly predictive when category i and 2 criteria are met.
REFERENCES 1.
Doust the
BD,
0:
Maklad
RR, Herman B-scan
68 : i 246-1
BB,
graphic
252,
Harris
11 :405-409,
J, Sarti Gastroen-
W: Ultrasonic a
Liver,
G,
comparison
gallbladder
and
radio-
Radiology
.
and
spleen.
Radiol
C/in
1975
Amberg
ultrasonic
ventional
MW, Castagna
1975
K, Brocker
13:543-55,
Leopold
scale
of
1974 EN:
NorthAm
examination 1974
ultrasonography.
cholecystography:
Carlsen
B-mode
1 1 1 :643-647,
of gallbladder
terology
5.
Ultrasonic
Radiology
J, Lindstrom
Value
3. Goldberg
4.
NF:
gallbladder.
2. Tabrisky
Discussion
The overall accuracy of ultrasound evaluation of gallstones has been previously reported, but without surgical confirmation in all cases and without evaluation of the accuracy of the individual categories of scanning patterns found. Our review of i45 surgically removed and pathologically studied gallbladders demonstrates that different categories of scanning images have different diagnostic importance. Shadowing opacities which move the gravity (category 1) were the most frequent findings in gallbladders containing stones, present in 92 of 137 abnormal cases (67%). Such findings are virtually pathognomic of gallstones. All 92 patients with this pattern had stones at surgery. The contracted gallbladder whose lumen is not seen on ultrasonic examination, even after fasting, has been previously reported. Leopold et al. [5] noted six cases,
229
CHOLECYSTOGRAPHY
J,
Gosink
BB,
cholecystography:
radiographic
Mittelstaedt
C:
a comparison
techniques.
Radiology
Gray-
with
con-
121 : 445-448,
1976 6.
Bartrum
R,
graphic
Crow
HC,
Foote
cholecystography.
SR: N
Ultrasonic
Eng/
J
and
Med
radio-
296:538-541,
1977 7. 8.
Monroe
SE,
der.
Med
Cal
Mujahed
bladder 9.
Baker
A, HL:
normal
1956
Evans
Further
JA,
Whalen
studies
on the
74:239-245,
SF: Performance
South
Mujahed
Congenital
Radiology
10. Ochsner 11.
FJ:
85:422-423,
absence JP:
The
on oral cholecystography.
tography. phy.
Ragen
Z:
Factors
gallbladder.
12.
Fischer
13.
cholangiography. Zboralske FE,
HW:
cholecystographic
interfering
Physiologic
accuracy
and
of oral
1974
cholecys-
of cholecystogra-
1970 Radiol
the
opacification 1:183-185,
pharmacologic
Radiol C/in North Am 4:623-625, Ainberg JR: Cholecysto-cholestasis: error.
gall-
112:1-3,
1960
with
Gastrointest
gallblad-
non-opacified
Radiology
and reliability
J 63:1268-1272,
Med
of the
Am J Dig Dis 7:339-346,
of
a
1976 aspects
of
1966 cause 1962
of
