Robert K. Zeman, MD #{149}William J. Davros, PhD #{149}Brian Steven C. Horii, MD #{149}Paul M. Silverman, MD #{149}Edward Wendelin S. Hayes, DO #{149}Cirrelda J. Cooper, MD
S Garra, MD #{149}Jo Anne L. Cattau, Jr, MD #{149}
Relationship between Stone Targeting, and Fragmentation during Experimental Biliary In vitro experiments in an anthropomorphic phantom were performed to clarify the relationship between stone motion, targeting, and fragmentation. Stone motion was minimized by pinning the stone against the dependent wall of a mock gallbladder cavity during shock wave treatment. Fragmentation was most effective (probably due to increased cavitation effects) when the shock wave traversed fluid at the point of its impact with a stone. The results suggest that treatment with the patient in the supine or oblique position may produce a better outcome than treatment in the prone position. Buoyant stones exhibited the greatest motion, which was often to-and-fro in nature. Although restricting the size of the mock gallbladder cavity reduced stone motion, maintaining a 1-cm fluid path was beneficial for achieving optimal pulverization. Index
terms:
Gallbladder, 762.1299
Radio!ogy
Gallbladder, interventional
#{149} Lithotripsy,
1990;
calculi, 762.289 procedures,
#{149}
762.1299 176:125-128
LTHOUGH
wave gational, cholelithiasis
been
with
fective targeting, fragmentation
from the patient with ment stones (patient The Lithostar Plus
shock
treated
this
otriptor
for
modality.
Ef-
stone motion, appear to have
and a corn-
was
used
system
has
electromagnetic
Medical to perform
three
pig-
Table).
shock
lith-
Systems,
Iselin,
lithotnipsy.
That
wave
generators:
under the treatment to biplane fluoroscopic
plex relationship, which gives rise to significant implications for the planning and performance of the lithotnipsy procedure. To elucidate the patterns of stone movement during lithotnipsy and how they may relate to targeting and fragmentation, in vitro ultrasonographic (US) and videofluonoscopic observations were made during fragmentation of stones
guidance.
The
third
overhead
module,
housed
ments, the stones were thropomorphic phantom phantom is made of an
in an anthropomorphic
MATERIALS A total patients
From the Departments of Radiology (R.K.Z., W.J.D., B.S.G., JAG., S.C.H., P.M.S., W.S.H., C.J.C.) and Medicine (Gastroenterology Division) (E.L.C.), Georgetown University Medical Center, 3800 Reservoir Rd NW, Washington, DC 20007. From the 1988 RSNA annual meeting. Received October 3, 1989; revision requested November 3; revision received March 12, 1990; accepted March 28. Address reprint requests to R.K.Z. RSNA, 1990
NJ)
(Siemens
predominantly
3 in the
Two are housed and are linked
phantom.
I
#{149}
Lithotripsy’
extracorponeal
have
RN
Motion,
lithotripsy is still investiseveral thousand patients
worldwide
Goldberg,
AND
of 35 stones at the
time
from
wave
the
11
of cholecystectomy
were
the
level
For the
shock were
properties
(and
-
m/sec
MHz.
therefore
phantom custom
had a diameter designed so that
dominantly
gallbladder
cavity
could
the
of the
phantom
all
came
of 1,000 expeni-
speed
1,540
shock to those
stones
generators
housed in an an(Fig 1). This agar and graphite
support similar
The
wave
fragmentation
were used in the fragmentation expeniments. The stones from 10 patients were predominantly composed of cholesterol, while those from one patient were prepigment.
at a
to 19 kV
A total
an approximate of dB/cm
used
given.
in vitro
transmission tion of 0.5
US
at full power, was
380 bar).
waves
an
in-line
equivalent
under-table
(approximately shock
used
module
power
in
uses
the purpose of these in the under-table shock
overhead
reduced on
For
generators
and
gel with
METHODS
harvested
for targeting. vitro experiments,
is contained
which
table
of sound
and
attenua-
Its acoustic its ability
wave transmission) of liver pamenchyma.
to are The
of 24 cm and a mock 50-mL
be carved
was
out
of
from single composition “families” (ie, had visually similar morphology and pigment content) and were of matched size (diameter, 1.0-1.6 cm). Prior baseline experiments had suggested that stones from
of the focal zone of the lithotriptor (1-3). Of the 35 stones treated, 31 were fragmented in the 50-mL mock gallbladder in normal saline or 7.5% or 15% urogmaphic
a single
contrast
composition
family
break
at com-
parable power levels with comparable degrees of pulverization. Between two and five matched stones were obtained from each patient. Three stones were obtained
interior
Abbreviation:
material.
Contrast
AP
at the
material
center
was
anteroposterior.
125
C
B
A
Figure Figure and
1. Treatment C an under-table
dent
or nondependent
tation
geometries shock wave
used in fragmentation generator (SWG)
position.
tangentially.
When
used,
experiments. used with
was
the US probe
With stones
tion size
geometries A in either a depen-
was positioned
to view
tiveness waves.
fragmen-
With
geometry B an overhead module was used for generating shock waves; stones were treated in the dependent position and were imaged with an in-line US probe. Of the 31 stones treated, 20 were treated with the shock wave traversing fluid to the point of stone impact (17 with geometry A, three with geometry B). Eleven stones were treated with geometry C. This does not include the stones treated in the smaller cavities.
administered
because
only
in biplane
to assist
geting
but
also
it could
be fluoroscopic
to manipulate
used
the
to buoyancy
not tar-
stone
position
due
effects.
tnibution ronment
of stones treated in each envifrom patients 1-11 is shown in
Table. Ten of the 1 1 stones trast material floated. contrast material, and
The
dis-
the
line.
Of
the
treated in 15% conFour floated in 7.5% three floated in sa-
17 floating
stones,
three
were
suspended in fluid and did not rise to the roof of the mock gallbladder. The pigment stones tended to become dependent in contrast
material.
graphic in two
evidence of the three
saline.
Stone
There
was
of gas stones
buoyancy
in
radio-
internal clefts that floated in
was
important
in
determining the position of a stone in me!ation to an incoming shock wave (Fig 1). Continuous
fluonoscopic
or
US
moni-
toring of the fragmentation process was performed and videotaped for review. The shock wave path was continuously readjusted so that the stone on largest dominant fragment was centered in the focal zone. If no dominant fragment was seen, the center of the mock gallbladder was targeted. Stone and fragment motion was rated on a crude motion scale: 0, no movement; 1, excursion of 1 cm or less; 2, brisk
motion.
The
actual
measurement
was made with use of an electronic grid, corrected for magnification and displayed on the viewing monitors. Motion was observed
with
genera! tion
each
trend. in
shock
The
relation
wave
direction
to the
and
mo-
surrounding
mcdi-
um and stone position were also noted. Measurements were performed 20 times each in the anteropostenior (AP) and oblique (craniocaudal) fluoroscopy planes or transverse
and
longitudinal
US
planes.
The stones and fragments were inspected at 200-shock wave intervals. At the conclusion
stone, 126
of fragmentation
the
fragments
Radiology
#{149}
were
for
each
segregated
and
photographed.
by
They
graded on a scale the effectiveness grade 2, excellent
mm,
abundant
were
3-5-mm
fragments);
0, poor fragmentation (failure stone breakage on fragments 80% of the original played little role mentation
matched
grade
Size frag-
stone
size were classification
5
to produce in excess of
stone diameter). determining
in
because
homogeneous jective, this
then
most fragments fragmentagreater than
used. does
sets
of
Albeit allow
sub-
crude grading of the effectiveness of fragmentation. The motion and fragmentation data treatment
will be presented separately geometries A, B, and C. The
Student significance
t test
the means indexes.
Seven
for motion
from
stone
and
additional
patients sets
for
was used to determine of the difference between
were
the
(from
four
or five
obtained)
were
volume
bladder mL
of the
cavity
or 6 mL.
With
was 1 cm or 1-2 in the near field
face for
with the
tro
mock
was
these
the stone. 15-mL
experiment
cavity.
Geometry The
conducted were
US
gall-
respectively, wave inter-
B was
in
used
visual
in the other repeated
15
there
and
in vi-
this
model.
None of the three stones treated with geometry B showed this effect, which occurred in only one of 1 1 stones
treated suggest
neal-time
with that
parallel
geometry this type
to the
C. These data of movement
direction
of shock
wave travel requires stone The three stones that were but
did
not
rise
nondependent
all
the
Off-center fragments
buoyancy. buoyant
way
surface
gallbladder showed tion. The rhomboid lanly shaped stones gyrate and tumble four instances.
but
to the
of the
mock
brisk level 2 moand more imnegucould be seen to end over end in
hits caused to be deflected
the stones laterally.
on
for 16 of 17 stones geometry A (Fig 3), one
treated
was not Treatment
for 17 stones motion index
duced 1.53
with
essential geometry
fluoro-
scopic and US guidance revealed that incoming shock waves caused stones and fragments to exhibit substantial motion within the mock gallbladder.
for
was
B, and
it. A was
used
and resulted in a mean of 1.88 ± 0.11. It pro-
a mean ± 0.26.
geometry
fragmentation For
this
ometry C. The mean was 0.55 ± 0.27, and
of
of stones
and very efFor the three
use of geometry index was 0.33 ± fragmentation in-
2.0 ± 0.0. This
lent fragmentation motion in this small stones were treated
index
group
there was brisk motion fective fragmentation.
dex of the
surface of the mock in 13 of 17 cases (Fig 3).
stones treated with B, the mean motion 0.33, and the mean
RESULTS Videotape
geometry but off the
of three
to either
mm of fluid, at the shock
shock
seven of 1 1 treated with geometry C. The extent of movement was included in the motion index calculations. Buoyancy favored lateral movement
fragment-
volumes
1,000
matched
elliptical
reduced
after
by
the four
ed in normal saline in a phantom identical to that already described except that the
nondependent gallbladder
This occurred treated with
fragmentation
stones
whom
of fragmentation
Directly centered hits with A caused little lateral motion caused the stone to ricochet
of 0-2 with regard to of fragmentation (Fig 2): fragmentation (no frag-
ments greater than 5 mm, 2-3 mm); grade 1, moderate tion (one to five fragments
observations
as a
of stone
size
2. Example of grade 1 fragmentaillustrates stratification of fragments for the purpose of determining effec-
meant
excel-
with little stone group. Eleven with use of ge-
motion index the mean fragJuly
1990
ThIGBL.m..
114GB 1.....
I
I
ri
I
i-.-
_________________________________________________
TM
GB
1..,,,,
,I
H15
a.
c.
b. 3.
vector representation of stone movement within tissue-mimicking mock gallbladder (TM GB). CC with stone recoil is greatest when an off-center hit drives the stone laterally within the AP viewing stones were also subject to random deflections off the nondependent wall of the mock gallbladder cavity. This movement the lateral deflections depicted in a. (c) Dependent stones treated with use of under-table shock wave generators exhibited than buoyant stones. This movement was in the general direction of shock wave propagation. Figure
Schematic
=
(a) Lateral movement
mentation index was 0.91 ± 0.09. This indicated moderate fragmentation with little to moderate stone movement. Student t tests were performed on motion and fragmentation mean indexes cant
to determine differences
whether signifiexisted between the
treatment geometries. For fragment motion, geometry A produced significantly greaten motion than geometries B and C (P < .001, t 6.74, 1.33).
stones
were
viewing
due
deflected
plane,
out
they
to partial
of the
were
volume
less
a second
ducen tudinal
and switching between and axial orientation,
were
able
out-of-line
to include
By
US
trans-
bongiwe
out-of-plane
motion in the index calculations. With both in-plane and out-of-plane motion, stones were seen to exhibit to-and-fro movement. After receiving an incoming shock wave, a stone
There was no significant difference in mean motion index between geometries B and C (P > .5, t = 0.635). For fragmentation, geometry B was
would be be drawn
found
all treatment geometries but was most pronounced with geometry A for buoyant stones. Recoil movement
to be marginally
superior
geometry
A (P < .5, t
try
superior
A was
1.54).
C in
producing fragmentation at the P .01 level (t 3.60), and geometry was superior to geometry C at the < .001 level (t 6.09). Five stones showed an unusual fragment spatial distribution that fected targeting. Two of the three
stones
that
internal
gas
floated
in saline
sank
to a dependent
effect
Geome-
to geometry
and
< B P
af-
had po-
sition immediately after being fnagmented. Radiognaphs of specimens of the fragments showed an absence of internal gas that had previously been present.
In
the
three
stones, the fragments ing out at different
remaining
were stratifylevels within the
mock gallbladder (one in 15% contnast material, one in 7.5% contrast material, and one in saline). Fragments from a single stone may have varying specific gravities due to their
composition, seek
different
which
causes
levels
within
them
to
than
with
US.
cross-sectional Volume
176
Because
nature Number
#{149}
of the
of US, when 1
was
was
present
position seen
(Fig
in all
in an
less
then
motion between
or fragmentation the results
mL cavities from treated
the
when
with
direction
for
Fragmentation effective
was in the
±
was in
was found in the 50- and
same composition with comparable 6-mL
15-
stones family geometry
considerably
B.
less
cavity.
DISCUSSION Targeting is a critical part of gallstone lithotripsy. This series of in vitro experiments sought to clarify the relationship between targeting, stone movement,
and
fragmentation.
motion
center
in the
stones
showed
wave
Al-
was
AP plane. the to the
B and
C un-
clearly
off
Buoyant
greatest
stones
rise
of stones
geometries
shock
that
motion.
did
not
corn-
nondependent
wall of the mock gallbladder showed the briskest motion and were most apt to migrate from the focal zone. Shock waves are longitudinal waves. Theoretically, stones could be propelled forward parallel to the dinection
of wave
propagation
during
the
compressive (positive-pressure) half cycle. This may account for the lack of stone motion with geometry B, wherein the stone is pinned against the mock gallbladder wall. The same logic would have predicted little motion with geometry A; however, this was not the case. Stones
the 0.33
index difference
comparing
the
pletely
12 stones and in a lateral direction for five. The limited experiments performed in a 6-mL cavity with each of four stones laterally surrounded by a thin 1-2-mm layer of saline revealed virtually no stone motion. A fragmentation index of 1 .25 ± 0.5 was achieved, meaning that fragmentation was effective but by no means
0.2, and the fragmentation 2.0 ± 0.0. No significant
with
Free-floating
and
complete. In the 15-mL cavity, mean stone motion index was
little
treated
3). This
media
axial
relatively
fluid.
This posed a challenge in targeting fragments separated from each other by several centimeters. Stone motion was better evaluated subjectively with biplane fluoroscopy
deflected and would back (often incompletely)
to its original
to
though US guidance is used for clinical lithotnipsy, we found that motion was more conveniently assessed in the laboratory with fluoroscopy than with US. Fluomoscopy allowed classification of complex motion because it is not cross-sectional in nature. The in vitro results demonstrated
visible
averaging.
using
craniocaudal. plane. (b) Buoyant was not as great as less movement
treated
showed the greatest ment, which, more ment
with
with
geometry
A
stone movethan the move-
geometries
B and
C, could
be described as to-and-fro. Although gravity may contribute to this peculiar motion in the axial direction, it is not likely to cause lateral recoil in a nondependent direction. There are three other possible reasons why during treatment stones may be dniven away from their original position and then pulled back. In addition to their compressive (positive-pressure) component, shock waves also have a ranefactive
ponent pressure localized
(negative-pressure)
(4,5). The component vacuum
corn-
latter
negativecould create effect, drawing Radiology
a
127
#{149}
stones
and
fragments
back
fragmentation should be maximized if stones can be kept in the shock wave focal zone by minimizing their
toward
the shock wave path after their initial deflection. Since the forces involved may be small, this effect may be obvi-
movement.
ous only for buoyant stones with melatively little inertia to overcome. A second possible explanation is that as negative pressures are created, cavitation bubbles are drawn out of solution (6,7). As these bubbles oscillate, are associated with jet formation, and ultimately collapse, they
may
exert
sufficient
buoyant
stones
A third shock
possible
waves
a cavity
may
nection ry wave.
force
to move
or small
fragments.
explanation
reflected
off
displace
stones
opposite to that We know that
are reflected just waves. Reflected
is that the
walls
of
in a di-
of the shock
primawaves
like other acoustic shock waves are not
uncommon and are responsible for most of the damage to in-line diagnostic US probes used in clinical lithotnipsy.
The vealed
in vitro experiments some unexpected
considerations. contain gas may release
ken.
Buoyant stones that in their internal fissures that gas after being bro-
Following
buoyancy sink into
below
the
concentrated
target
and latter
of this
gas,
may
in a stone
zone.
Tar-
complicated according Most stones
pigment be
rim
cornfocally
or nidus
(8). If that is the case, fragments from that area may seek greater dependency than those with greater relative cholesterol content. Aiming at fragments distributed oven several centimeters is problematic. They generalby must be treated separately unless they can be made to lie in the long axis of the shock wave focal zone. It might be assumed that stone
128
Radiology
#{149}
suit
in
This
study.
was
not
Geometry
fragmentation
strictly C did
as
true not
me-
effective
as
that
with geometries A and B. GeomA resulted in good fragmentation despite much stone motion. Stones treated with geometry B (although a small group) showed excellent fragmentation with little movement. Comparable results were obtamed in the l5-mL cavity with use of this same geometry. These data suggest that shock waves produce more effective fragmentation if they traverse fluid up to the point of impact with a stone. This may indicate that a supine on oblique treatment position may result in better fragmentation than a prone treatment position with stones resting on the dependent gallbladder wall. The mesults of the limited fragmentation experiments conducted in the 6- and l5-mL cavities suggest that fragmentation can occur if stones are bathed in a small amount of fluid (1-mm fluid interface); greater quantities of surrounding fluid (i-cm fluid inter-
etry
however,
to
cabby contracting the gallbladder reduce stone motion, but if the bladder is contracted too much, effectiveness of fragmentation be compromised. The presence surrounding fluid may promote tation effects, which are known participate in stone fragmentation
stones
move
in the
exposed ometry
surface can
of a stone.
be achieved
This while
be
authors
frag-
thank
Yvonne
preparation
of this
References 1.
Davros
WJ,
Zeman
mimicking the
effects
(abstr). 2.
Choyke
PL,
Pahira
an in vitro 170:39-44. man
BS, Davros
using
4. 5.
7.
8.
to 9.
Davros
WJ,
Nilges
E,
Radiology
WJ, Lack
1989;
EE, Horii
of gallstone and
SC, Ze-
fragments
fluoroscopy:
implica-
tions for monitoring of gallstone lithotripsy. Radiology 1990; 174:343-347. Reichenberger H. Lithotripter systems. Proc IEEE 1988; 76:1236-1246. Coleman AJ, Saunders JE. A survey of the acoustic
6.
gallstones
Renal calculi after US evaluation with
Visibility
ultrasound
Tissue-
l69(P):380.
JH,
phantom.
RK.
BS.
in studying
on
1988;
AJ, Mun 5K. wave lithotripsy:
Garra
Garra
for use
of lithotripsy
Radiology
Dwyer shock
3.
RK,
phantom
output shock Bio!
of commercial wave
1989;
extracor-
lithotripters.
Ultrasound
15:213-227.
Ter Haar G, Daniels 5, Eastaugh KC, Hill CR. Ultrasonically induced cavitation in vivo. Br J Cancer 1982; 45(suppl):l51-i55. Coleman AJ, Saunders JE, Crum LA, Dyson M. Acoustic cavitation generated by an cxtracorporeal shockwave lithotripter. Ultrasound Med Biol 1987; 13:69-76. Trotman BW, Morris TA, Sanchez HM, So!oway RD. Ostrow JD. Pigment versus cholesterol
due
path
effective
The
Med
the impact of shock waves and often recoil. This type of movement is most common for buoyant stones. Fragmentation is most effective when shock waves traverse fluid up to the
fluid
Acknowledgment:
poreal
may gallthe may of cavito
a i-cm
to produce U
Carew for assistance manuscript.
(9). In conclusion,
that
maintained mentation.
face) allow better fragment pulvemization. Theoretically, pharmacobogi-
and fragments position well
be further of fragments gravity.
cholesterol
ponents;
escape
original
geting may by layering their specific
have
the
is reduced a dependent
the
also metargeting
our
in
same time limiting movement of nonbuoyant stones by pinning them against the dependent surface of the gallbladder. Limiting the amount of fluid surrounding a stone (as would occur with pharmacologic contraction of the gallbladder) can reduce stone movement. It is recommended,
cholelithiasis:
identification
and quantification py. Gastroenterology Delius MK, Brendel tion in extracorporeal
by infrared spectrosco1977; 72:495-498. W. Mechanisms of acshockwave lithotrip-
sy.
Delius
In:
Ferrucci
MJ, eds. Biliary Book Medical,
JT,
lithotripsy. 1989; 31-42.
MK,
Burhenne
Chicago:
Year
geat the
July
1990