Kazuhiro Tsuneo
Shimamoto, MD #{149} Sadayuki Ishiguchi, MD #{149} Shigeki Itoh,
Hepatocellular with Color
terms:
Blood,
flow dynamics,
761.321
Blood vessels, MR studies, 952.1299 #{149} Blood vessels, US studies, 952.1298 #{149} Liver neoplasms, blood supply, 761.321 #{149} Liver neoplasms, CT, 761.1211 #{149} Liven neoplasms, MR studies, 761.1214 #{149} Magnetic resonance (MR), companative studies #{149} Magnetic resonance (MR), vascuIan studies, 761.1299 #{149} Ultrasound (US), companative studies #{149} Ultrasound (US), Doppler studies
Radiology
Takeo Fukatsu,
1992; 182:149-153
C
Ishigaki, MD
#{149}
Carcinoma: Doppler US and
Fifteen patients with hepatoceblular carcinoma underwent examination with color Doppler ultrasound (US), magnetic resonance (MR) imaging, dynamic computed tomography (CT), and angiography. Doppler signals ranging from 0.22 to 3.48 kHz could be obtained from within the tumor in 14 of 15 patients. The resistive index was 0.38-0.77. Color Doppler signals were visualized in nine of 15 lesions with a Doppler shift greater than 0.7 kHz. The Doppler signals and the resistive index of the tumor vessels became lower as the vessels progressed into the center of the lesion. The appearance of tumor vessels recognized on MR images obtained with gradient-recalled acquisition in the steady state (GRASS) in 11 of 15 besions was compared with that on CT scans and angiograms. Tumors that were hyperintense on GRASS images obtained with a flip angle of 15#{176} transmitted Doppler signals of considerably higher amplitude compared with the isointense lesions. Both color Doppler US and MR imaging provided useful information for characterizing intratumoral blood flow. Index
Sakuma, MD MD #{149} Hiroshi
MD
Evaluation MR Imaging’
Doppler ultrasound (US) has benefits in evaluation of the hemodynamic characteristics of abdominab tumors (1-3). The flow distnibution within the tumor and analysis of Doppler spectral patterns are helpful for the differential diagnosis of hepatic tumors and coincide with the gray-scale US findings (2,3). Although technical problems still exist in detection of slow flow in the lesions, the latest colon Doppler equipment can show characteristic flow patterns in. 80%-95% of hepatocellubar carcinomas (HCCs) (2,3). On the other hand, magnetic resonance (MR) imaging is also an estabbished diagnostic modality for detection of focal liven lesions and characOLOR
temization HCC
of the internal
(4,5).
To our
structure
knowledge,
ever, the correlation
of
how-
between
signal
intensity patterns and tumor vasculanity has not been evaluated. This study was designed to explore the characteristics of blood flow in HCC with the help of Doppler spectrab analysis and to correlate colon flow patterns with MR images and dynamic computed tomographic (CT) scans. We usefulness
uation
also studied the of MR imaging
of intratumonab
potential in the evab-
blood
flow.
MATERIALS Fifteen
AND
patients
circumscribed)
29,
1991;
received
reprint
requests
RSNA,
1992
to 5.5.
seven
patients,
typical clinical
dy-
hypervasdiagnosis
of
angiography.
An-
giography was performed in the preembolization wonkup. A commercially available color Doppler system (SSD-680; Aloka,
Tokyo)
and a convex
insonating
transducer
frequency
used in this study. flow velocity was ten of 100 Hz was
of 3.5
The
the
were
minimal
2 cm/sec. used.
with
MHz
detectable
A Doppler
At color
fib-
flow
imag-
red indicated flow toward the transducer, and blue, flow away from it. Blood flow detected within the tumor was coning,
sidered
abnormal
vessels,
and
with
flow
within
its distribution
dynamic
CT
scans,
diseased was
the corresponding
compared
distribution MR
grams. For quantitative color flow image was
images,
on and
angio-
Doppler study, the used as a guide to
obtain the Doppler spectrum. When colon Doppler signals were detected, periphery of the lesion was examined with the maximum sampling volume
no the (10
mm) to search for abnormal signals. To avoid errors in flow classification between
tumor
vessels
and the surrounding
patic arteries that did the sampling volume mm for recording the examination could be
freezing help
the B-mode
of breath
normal
signals with
selected
with
METHODS the
or massive
nodular
Doppler
US, MR imaging,
(well-
(ill-defined)
type of HCC (6,7) (12 men and women aged 47-87 years [mean years]) underwent examination
tamed the Department of Radiology, Nagoya School of Medicine, 65 Tsunuma-cho, Nagoya 466, Japan. Received April revision requested June 24; revision July 30; accepted August 12. Address
In these
the
colon image.
holding,
several
could
be identified.
highest
for recording
he-
not supply tumors, was reduced to 1-2 signals. The Doppler performed without
With sites
frequency
the
the
of ab-
The shift
was
time-velocity
waveform.
3.8 From University Showa-ku,
liver).
namic CT demonstrated cular patterns (8), and HCC was made before
point
three age,
61
with
dynamic
colon
CT,
and angiography at Nagoya University Hospital, Nagoya, Japan. The lesions were 1.0-12.0 cm in diameter (average diameter,
I
in the
cm).
The by
pathologic
means
of surgical
proof
was resection
fine-needle biopsy in eight patients. seven other patients, the diagnosis
ob-
of primary
malignant
tumors
evaluate
(9). In addition,
the
Doppler waveform recorded at the niphery of the tumor was compared that recorded at the center.
impedance
pewith
or
In the was
established with the help of clinical data and imaging studies (liver cirrhosis, ebevated serum a-fetoprotein level, and absence
The peak systolic frequency shift was used to estimate flow in tumor vessels. Angle correction was not used because in most cases a large enough portion of the tumor vessels was not visualized. The resistive index, (A - B)/A (where A is the peak systolic frequency shift and B the end-diastolic frequency shift), was used to
except
Abbreviations:
AV = anteriovenous, GRASS gradient-recalled acquisition in the steady state, HCC = hepatocellular carcinoma, SE = spin
=
echo.
149
MR imaging
was
superconducting Systems, section
performed unit
with
(Signa;
Milwaukee). axial spin-echo
GE
a 1.5-T
Medical
Initially, the (SE) images
multiof the
entire liver were obtained with an acquisition matrix of 256 x 128, section thickness of 7 mm, and intersection gap of 3 mm. A pulse sequence of 500/30 (repetition time msec/echo time msec) was used to obtain
TI-weighted 2,000/80 images.
were
images;
was used Subsequently,
obtained
sequence
acquisition
state
(GRASS).
ages
were
obtained:
70/28
with
with
8
a flip
a flip
9 10
angle
11 12 13 14 15
evaluate
analysis,
was
applied
of the signal
to Doppler index.
f (Id-Iz)
Signalst
8 4 5 3 5 5 8 1 5 5 6 8 2 8 8
Evaluation
Doppler
US, MR Imaging, Modality or Criterion
Color
in-
frequency
shift
RESULTS
-
0.22 0.45 0.48 0.69 0.69
+
0.74
+ +
0.81 0.96 1.00 1.88 2.68 2.73 2.96 3.48
-
+ + + + + +
ysis
are
summarized
in Tables
US
absent,
=
the
mass
along
2 and
the
picted in nine of 15 lesions with a Doppler shift greaten than 0.7 kHz. No lesion had a frequency shift exceeding 3.5 kHz. Doppler signals could not be obtained from one lesion 3 cm in diameter located in the anteriom superior segment (patient 1), abthough it was hypervascular on both dynamic CT scans and angiograms. The resistive index was 0.38-0.77 (mean, 0.57 ± 0.11). As shown in Table 2, the peak systobic frequency shifts in Doppler signabs from HCCs greaten than 2 cm in diameter (mean, 1.69 kHz ± 1.08) were significantly higher than those from tumors less than 2 cm in diame#{149} Radiology
-
+ -
-
-
+
+
-
+
+
+ +
+ +
+ +
+
+
+
+
+
+
+
+
+
+
+
+
-
In patient
1, Doppler
signals
could
not be
of Blood and
Flow Depending Dynamic
on Vascularity
of Tumor
at Color
CT Quantitative No. of Cases
Doppler
M (kHz)
Data
Resistive
Index
MR imaging
-
9 4 10 2
+
12