AJR 2000; 174:834-836
© American Roentgen Ray Society
MR-Guided Biopsy Using Respiratory-Triggered High-Resolution T2-Weighted Sequences
H.-J. Langen1,2,
H. Kugel1,
S. Grewe1,
A. Gindele1,
P. Landwehr1 and
R. Fischbach1
1
Department of Diagnostic Radiology, University of Cologne, Joseph-Stelzmann
Str. 9, D-50924 Köln, Germany.
2
Present address: Department of Diagnostic Radiology, Missionsaerztliche
Klinik, Salvatorstr. 7, 97074 Wuerzburg, Germany
Received January 25, 1999;
accepted after revision August 9, 1999.
Address correspondence to H.-J. Langen.
Introduction
The use of MR imaging for imaging-guided biopsies can be the method of
choice for certain patients. Because of the high soft-tissue contrast, some
lesions can be visualized more clearly—sometimes exclusively—with
MR imaging. In addition, slice orientations, allowing access routes that are
difficult to plan and to perform with CT, are freely selectable
[1,
2]. A further advantage is the
lack of radiation with MR-guided biopsies. They became possible with the
introduction of nonferromagnetic MR-compatible biopsy needles
[3].
For MR-guided biopsies in the abdomen, pulse sequences with short
acquisition times are usually used to obtain MR images during suspended
respiration. Compared with standard diagnostic pulse sequences, they suffer
from lower spatial resolution
[2] and have inferior contrast
and lower signal-to-noise ratio
[4]. Also, these sequences are
more sensitive to susceptibility artifacts
[4]. In addition, similar to
CT-guided biopsies, the degree of the patient's inspiration may vary in
sequential breath-holds; thus, the anatomic relations may change between
consecutive control measurements. To improve the image quality during
MR-guided biopsy and to avoid the unreliable uncomfortable breath-hold
technique, we used a non—breath-holding T2-weighted imaging sequence
triggered by the spontaneous respiration of the patient.
Materials and Methods
Three patients (two women, one man), 58, 66, and 69 years old, underwent
MR-guided biopsy with a 1.5-T MR scanner (Gyroscan ACS NT; Philips Medical
Systems, Best, the Netherlands). All three patients had suspicious liver
lesions diagnosed on MR imaging. The mean diameter of the lesions was 6.2 cm
(range, 2.5-10.0 cm); the mean distance of the access route was 3.7 cm (range,
2.0-6.0 cm). Two lesions were biopsied twice with 18-gauge biopsy needles; the
third lesion was biopsied once with a 14-gauge needle. One patient was deaf
and did not react to instructions for breath-hold.
A grid (constructed locally) was used to plan the access route to the
lesion. The grid was made up of seven polyethylene tubes (length, 10 cm;
diameter, 5 mm). The tubes were filled with water; each end was fused over an
open flame. Water was chosen to produce a high signal intensity on the
T2-weighted images. The tubes were placed on adhesive tape at 1-cm intervals
to form a grid.
Before the intervention, an 18-gauge venous cannula was placed in an
antecubital vein to allow immediate access in case of complications. Pulse
oximetry was fixed on a finger and a pneumatic respiration sensor—part
of the standard system equipment—was attached to the patient's abdomen
with a belt. On the basis of previous MR images, the estimated location of the
cutaneous entry point for the biopsy needle was marked on the patient's skin
with a pen; this mark was then aligned with the transverse plane of the
isocenter of the magnet with the scanner's light visor. At this location, the
grid was placed on the body surface, with the tubes oriented perpendicularly
to the planned transverse imaging slice (Fig.
1A,1B).
View larger version (171K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A. —69-year-old woman with hepatic metastasis from adenocarcinoma.
Respiratory-triggered T2-weighted MR image shows biopsy planning of hepatic
lesion. Tubes of grid (arrow) are positioned orthogonal to imaging
plane that is perpendicular to body surface and encompasses planned access
route. After localizing target lesion, possible access path can be followed in
full length on this image.
|
|
A multislice survey scan was performed with the following parameters:
coronal, sagittal, and transverse orientation; turbo field echo; TR/TE,
6.8/3.4 msec; flip angle, 25°; field of view, 450 x 450 mm; slice
thickness, 15 mm; matrix size, 256 x 128; duration, 9 sec. On the basis
of this survey scan, a transverse respiratory-triggered T2-weighted turbo
spin-echo sequence was defined for planning the access path and for all
consecutive control measurements. The acquisition was divided into a number of
pulse trains triggered when expiration was signaled from the respiration
sensor. Because the pulse trains lasted 1.8 sec, they were finished before
inspiration. The actual TR is given as the interval between consecutive
inspirations. The nominal TR of 1800 msec is shorter than this interval and
gives the lower limit of scan repetition. Other scan parameters were the
following: TE, 93 msec; field of view, 345 x 242 mm; number of slices,
nine; slice thickness, 8 mm; number of excitations, one; echo train length,
21; echo spacing, 8.5 msec; acquired matrix, 256 x 125; rectangular
pixels, 1.35 x 1.94 mm. Scan duration was 50-60 sec, depending on actual
length of respiratory cycle. Increasing actual TR over the nominal value of
1.8 sec has only negligible influence on the T2-weighted image contrast.
After localizing the target lesion on the MR image, the most convenient
access path was determined on one image slice. The craniocaudal position of
the entry point was calculated as the distance from the MR slice showing the
access route to the slice through the magnent's isocenter, indicated by the
mark on the skin. With the grid on the slice showing the access route, we
determined the left-right coordinates of the entry point. To check the planned
path, we partially moved the patient out of the magnet bore, and a marker with
high signal intensity on T2-weighted images (e.g., a nifedipine capsule) was
fixed at the planned entry point. Then the patient was moved into the magnet
bore again to the identical position, with a standard deviation of ± 2
mm, to obtain a control image with the marked entry point.
The patient was moved out of the magnet bore until the entry point was
freely accessible; then the biopsy was performed. After local anaesthesia with
10 ml mepivacain 1% (Scandicain 1%; Astra Chemicals, Wedel, Germany), a small
incision was made with a stylet. For biopsy we used MR-compatible biopsy
needles: 18-gauge, 100 mm (MRT-Biopsie-Handy; Somatex, Berlin, Germany) and
14-gauge, 100 mm (Biopsy-Gun; Daum, Schwerin, Germany). Metallic biopsy
needles can be seen from the signal void they produce. With the applied turbo
spin-echo sequences, these needles showed a signal void of less than 9 mm
diameter, which was sufficient to detect the needles in an environment with
high signal intensity without masking important structures
[5]
(Fig. 1B).
The biopsy needle was pushed forward directly into the lesion or until the
needle was stabilized in the soft tissue. Images obtained using respiratory
triggering were equivalent to a view during breath-hold in expiration;
therefore, the needle was advanced only at expiration. The position of the
needle tip was checked in the magnet; then the needle was advanced into the
lesion. A new control image was obtained to prove that the needle was adjacent
to the lesion. After moving the patient out again, we released the biopsy gun.
The needle position in the lesion was revealed by MR imaging, and the biopsy
needle with the specimen was removed. Using the same entry point, we performed
a second biopsy on two patients. A final MR image was then obtained to exclude
early complications. The total duration of the biopsy protocol was calculated
from the time stamps on the MR images. The time for biopsy was defined as the
period from the first survey to the last sequence (MR control image after
biopsy). For the total time the scanner is occupied, the in-room setup time
must be added.
Results
In all three patients, the liver lesions could be exactly delineated on the
respiratory-triggered T2-weighted sequences. Motion artifacts were not
observed. The position of the biopsy needle could be accurately identified in
all three patients (Fig. 1B). In all patients, the biopsy specimens were sufficient for histologic
evaluation (two cholangiocellular carcinomas and one metastasis from an
adenocarcinoma). The mean duration of biopsy was 35 min (range, 31-37
min).
Discussion
Conventional short breath-hold pulse sequences have the disadvantage of
inferior image quality because of low spatial resolution, inferior contrast,
and low signal-to-noise ratio
[2,
4]. In our opinion, this
disadvantage offsets the advantages of MR imaging, especially in
interventional procedures. A respiratory-triggered technique allows higher
quality MR images and obviates unreliable and inconvenient breathholds by the
patient. In addition, the application of spin-echo instead of gradient-echo
sequences reduces the size of the signal void produced by the metallic needles
and thus allows a more exact determination of needle position. In spite of the
longer image acquisition time with our protocol, the duration of the biopsy
procedure is within an acceptable range. The procedure permitted a controlled
biopsy with an exact identification of tumor and needle position, and we were
able to procure sufficient biopsy specimens in all three of our patients.
Our MR-guided biopsies were performed with a standard 1.5-T MR scanner,
which is not specially designed or equipped for interventions; thus, our
protocol should be transferable to other MR systems. However, depending on
specific features of the scanner, modifications may be necessary. On lower
field systems the signal-to-noise ratio may be reduced. This problem can be
overcome by increasing the number of acquisitions. The increasing scan
duration should not pose a problem using respiratory-triggered sequences.
Also, the size of the signal void may become too small to allow a reliable
identification of the needle. To ensure visibility of the needles in these
cases, the radiologist may enlarge the signal void with respiratory-triggered
gradient-echo sequences or needles of different alloys
[5,
6].
The diameter of the bore in our MR system is 42 cm (anterior-posterior) and
60 cm (left-right). We had no problems rescanning the patients after insertion
of the biopsy gun. However, problems rescanning patients after insertion of
the biopsy gun may occur in smaller magnet bores if the patient is too large
or if 150-mm biopsy needles are required. For these patients, a coaxial
system, 16-gauge cannula with a 3-cm handle (MRI Pencil Tip Puncture Needle;
Somatex) is available.
In open MR systems, biopsy devices may be handled a different way. The
specific advantages of respiratory-triggered T2-weighted turbo spinecho
sequences are not affected. The problem of biopsies in patients who are not
able to perform reproducible breath-holds may be solved using cine CT or
respiratory-triggered CT. However, experience with these methods is limited
[7,
8]. The determination of the
skin entry point was facilitated by our marker grid. If T1-weighted sequences
are necessary, a similar approach can be used with a marker grid providing
high signal intensity on T1-weighted images. A marker grid (Targogrid; Daum)
has become commercially available recently.
We believe that our technique extends the indications for MR-guided
biopsies. In the past these indications have included biopsy of lesions
detectable only on MR imaging and lesions with complicated angulated
approaches [2]. Our technique,
using respiratory-triggered sequences, allows biopsy of patients who cannot do
reproducible breath-holds. This technique may offer an additional advantage
compared with CT-guided biopsies. Further investigations in larger groups of
patients are necessary to confirm these findings.
References
-
Buecker A, Adam G, Neuerburg JM, Glowinski A, van Vaals JJ,
Guenther RW. MR-guided biopsy using a T2-weighted single-shot zoom imaging
sequence (Local Look technique). J Magn Reson Imaging
1998;8:955-959[Medline]
-
Gehl HB, Frahm C. MRI-controlled biopsies [in German].
Radiologe
1998;38:194-199[Medline]
-
Lufkin R, Teresi L, Hanafee W. New needle for MR-guided aspiration
cytology of the head and neck. AJR
1987;149:380-382[Free Full Text]
-
Kahn T, Schwarzmaier HJ, Modder U. Interventional MR tomography:
current status and future perspectives [in German]. Radiologe
1998;38:159-167[Medline]
-
Langen HJ, Kugel H, Heindel W, Krahe T, Gieseke J, Lackner K.
Localization of puncture needles in MRI: experimental studies on precision
using spin-echo sequences at 1.0 T [in German]. Rofo Fortschr Geb
Roentgenstr Neun Bildgeb Verfahr
1997;167:501-508
-
Lewin JS, Duerk JL, Jain VR, Petersilge CA, Chao CP, Haaga JR.
Needle localization in MR-guided biopsy and aspiration: effects of field
strength, sequence design, and magnetic field orientation. AJR
1996;166:1337-1345[Abstract/Free Full Text]
-
Katada K, Kato R, Anno H, et al. Guidance with real-time CT
fluoroscopy: early clinical experience. Radiology
1996;200:851-856[Abstract/Free Full Text]
-
Mori M, Murata K, Takahashi M, et al. Accurate contiguous sections
without breath-holding on chest CT: value of respiratory gating and ultrafast
CT. AJR
1994;162:1057-1062[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Introduction
Materials and Methods
