Therapeutic Soheil L. Hanna, H. Logan Brooks,
MD #{149} Michael A. Lemmi, MD Jr, MD #{149} Suzanne Gronemeyer,
Treatment MR Imaging of Radioactive
terms:
Eye, MR studies,
224.1214
neoplasms, gy, 22.1299
224.371 #{149} Eye. therapeutic #{149} Melanoma, 224.371
Radiology
1990; 176:851-853
Eye,
#{149}
radiolo-
B
125 for
used
noma quimes ing
gold plaque treatment
(1,2). Optimum accurate design
of the
1-125
perimeter margins calculation tion
ing
(S.L.H.,
Oncology Hospital,
Departments M.A.L.,
(J.F.), 332
J.W.L.,
ship
the
St Jude N Lauderdale,
Imag-
and
Children’s
Radiation
1990;
revision
requested
April
Fontanesi,
MD
its
tumor
may
affect
as improper from
base
(3).
this
relationship
loose
of the
and
mecti muscles. position, ei-
tion
obtained
B-scan
tumors
with
as 3 mm
that have well-defined gins. However, tumors distinct
margins
(US)
or those
for
marflat in-
that
the
gold
efficacy ing the
MR
plaque
extend
appearance
and
of MR imaging plaque/tumor
15 consecutive
of the
determine
the
in establishrelationship
in
cases.
revision
received and accepted May 1. Supported in part by the National Cancer Institute Cancer Center Support (CORE) grant P30 CA 21765 and the American Lebanese Syrian Associated Charities (ALSAC). Address reprint requests to J.W.L. RSNA, 1990
Plaque margin
MATERIALS
AND
The plaques
used
of the
radio-frequency
tion
on
as to ensure
that
a uniform
dose
were
a manner
rate
tumor
magnetic
a
field
(RF) power tested
a high-duty
5
of ophthalmoscopic In vitro determina-
of the
was
over
by
deposi-
placing
a gold
at room temby MR imaging
cycle
RF power
se-
quence. No softening of the margarine, which was determined to melt at 28#{176}C, resulted, and no plaque motion occurred. Episclerab plaque placement was performed in 15 patients with choroidal melanoma (mean age, 65.9 years). The plaques ranged in diameter from 8 to 19 mm
and
were
margins with the boundaries turing
the
placed
by
defining
transillumination, on the scbera,
plaque
tumor
marking and then su-
accordingly.
cation and tumor with MR imaging
coverage within
Plaque
were 2 days
bo-
assessed of sum-
gery. MR imaging unit
was
(Magnetom;
performed
Siemens,
on a 1.0-T Iselin,
NJ)
with use of a linearly polarized head coil. Routine evaluation included Ti -weighted sequences (repetition time [TR] was 550 msec, echo time [TE] was 17 msec [TR/TE 550/
17],
pentaacetic echo time,
in such
effect
gold
spread
slots
of the
chosen to provide the tumor base as
on top of margarine (20#{176}C) followed
(2,500/25
the
custom-
and size
to the
size was around
tion
customized according to the shape and size of the tumor as determined at US (Fig 1). Plaques consisted of a concave gold shell with a Silastic insert (Dow-Corning, Midland, Mich) containing slots in which 1-125 seeds were placed. The seeds were within
plaques
two
signals
averaged,
and
field of view was 25 cm) in at least two orthogonal planes. Oblique planar imaging was rarely used. T2-weighted images
METHODS
in our study
delivered
and
=
TN
16;
was
days. 2-mm
with
in height
sharp with
cGy/h
plaque perature
high-mesolu-
ubtmasonogmaphy
as small
gold
to the shape
determined by means and US examination.
them at surgery or in the early postopenative period, is desirable to enable plaque repositioning or dose mecalculation (3). Plaque/ tumor relationship can be readily
Concave
Various
sutures,
adjacent of plaque
1.
ized according tumor.
relation-
estimation
diameter,
pressure Verification
this
Figure
epis-
Research
Memphis,
38101, and the Departments of Radiology (S.L.H., M.A.L., J.W.L., J.F., S.C.) and Ophthalmology (H.L.B.), University of Tennessee College of Medicine, Memphis. From the 1989 RSNA scientific assembly. Received February 12,
that
placed
maintains
to the
tumor
such
surgically
plaque
factors such
plaque
anterior to the oma semmata are difficult to assess with US (4). Although the diagnosis of choroidab melanoma with magnetic resonance (MR) imaging has been reported (5.-il), MR evaluation of the 1-125 plaque/tumor relationship has not been fully evaluated (12). The purpose of this study is
of Diagnostic 5G.)
#{149} James
treatment meand position-
extends 2 mm beyond the of the tumor base (2). Dose is based on the assump-
that
clemal
1-125 the
MD
with an iodineis commonly of choroidal mela-
RACHYTHERAPY
to describe
t From
W. Langston,
PhD
of Choroidal Melanoma: in the Assessment Plaque Position’
Verification of the position of an episcieral iodine-125 gold plaque in relation to underlying choroidal melanoma is essential during early radiation therapy to ensure accurate plaque placement and thus optimum dose delivery. The authors used magnetic resonance (MR) imaging to examine 15 patients with choroidal melanoma after plaque placement to assess tumor coverage. The relationship of the plaque to the tumor was well defined in all cases, including two tumors anterior to the ora serrata. MR imaging measurements of the plaques were within 1 mm of the actual plaque sizes, while tumor measurements were within 2 mm of the preoperative ultrasound estimations of tumor dimensions. Tumors as small as 3 mm thick were readily visualized with MR imaging. Associated subretinal effusion was demonstrated in seven cases. Index
#{149} James
Radiology
of 65
and
90,
one
signal
averaged,
and field of view of 25 cm) were then tamed in the most appropriate plane.
Abbreviations:
DTPA
obA
diethylenetniamine-
acid, RF radio frequency, TR = repetition time.
TE
851
256
X 256
matrix
of 5 mm
were
with
used
were
thickness
intersection
gap
Ti-weighted
repeated
intravenous
linium
a section
in all sequences.
sequences
the
and a 1-mm
in
12 cases
administration
after
of gado-
diethylenetriaminepentaacetic
acid
(Magnevist; Berlex Imaging, Wayne, NJ) at a dose of 0.1 mmol/kg body weight. The following data were then recorded
for
sion,
each
patient:
presence
and
hancement, or undetected), position
subretinal plaque
(loose
or
dimen-
of tumor
effusion diameter,
tightly
tumor relationship tric, or malpositioned),
nerve
tumor
pattern
(well
en-
(detected plaque
placed),
plaque!
centered, eccenand plaque/optic
relationship.
RESULTS
a.
b.
Figure The
gold
plaque
ed on MR absent
was
images
signal
well
delineat-
as a curved
adjacent
strip
to the
of
outer
2.
showing applied
Sagittal
(a) and coronal
the plaque against the
as a curved bright tumor,
(b) Gd-DTPA-enhanced
Ti-weighted
images
(550/17)
dark signal void (arrow). Plaque is well centered which is contrasted between dark plaque and
and tightly vitreous fluid.
as-
pect of the sclema (Figs 2, 3). Contrast between the plaque and tumor was best and
depicted on the Gd-DTPA-enhanced
ed images
(Fig
4). Plaque-tumor
tmast was maintained hanced Ti-weighted trast between the
was
decreased
proton-density Ti-weight-
con-
on the images tumor and
in one
nonenbut convitreous
lesion
consis-
tent with the low melanin content of the tumor (6). Fourteen plaques were found to be tightly applied to the outer sclenal surface, while one was
seen
with
Plaque
its posterior
measurements
were
within
dimensions. demonstrated ly differentiated retinal effusions
seven
cases
edge on
raised.
MR
±i mm of their The tumor was
images
actual well
in all patients and easifrom adjacent subthat were present in
(Fig
5). Tumor
measure-
ments with MR imaging were within 2 mm of their estimated size with US. Lesions as small as 3 mm thick were demonstrated with MR imaging. Images from 10 of the 12 patients who underwent Gd-DTPA-enhanced MR imaging showed homogeneous lesion enhancement. Two lesions were not enhanced. The relationship of the plaque to underlying tumor was well defined in all 15 cases, including the two
(13%)
in which
the lesions
extended
anterior to the oma semmata (Fig 6). Fourteen of the 15 tumors were completely covered by the plaque. In 12 of those i4 cases, the plaque was well centered with its edges extending beyond the tumor boundaries in all directions (Figs 2, 3). The other two
were
found
eccentrically
with at least tumor margin the 15 tumors
emed (Fig 852
6). The
#{149} Radiology
positioned
one edge abutting the (Fig 4). Only one of was partially uncov-
relationship
of the
a.
b.
Figure
3.
Sagittal
Ti-weighted
(550/i7)(a)
and T2-weighted
a well-centered
plaque (arrow) completely covering two components of different brightness consistent (b) Contrast between the plaque and tumor is lost
plaques to the optic nerve was well demonstrated in all cases. The edges of the plaques were separate from the optic nerve in all cases except in one in which theme was subsequent plaque tilt. Adequate imaging was obtained in 13 of the 15 cases with use of orthogonal planes. Oblique planar imaging was necessary in the remaining two patients in whom the long axis of the tumor was obliquely oriented.
Optimal
brachytherapy
for
melanomas cleral plaque
requires that be accurately
against
tumor
the
the
calculated
the
tumor
base
dosage (2,3).
the
choroidal
1-125 placed
to ensure
is delivered
Verification
of
episthat
to plaque
images
showing
or repositioning or dose recalculation (3). When available, intmaoperative US evaluation of the plaque position is the position
is essential
soon
thereafter
simplest
means has
prior
to wound
of
been
to
optimal
(4).
closure
the
ora
absence MR
assessment,
of
its
metallic
objects,
promise
CT images.
short
Ti
but
also
serrata
This helps
and
be
US chosen
T2 not
ca-
tolerance
of
seriously
Also,
corn-
choroidal
characteristic
the
sub-
imaging
which
of and
may
greater
show
because
melanin.
its
the
place
(3). However, that extend
multiplanar
and
nals
after in
of intraoperative imaging was
pabilities
melanornas
surgery
verification
of tumors
anterior
at
plaque
sutured
US evaluation
In the
either
to permit
plaque
because
DISCUSSION
(2,500/90)(b)
the overlying lesion. (a) Tumor shows with varying melanotic concentrations. on the T2-weighted image.
inherent,
relaxation only
differentiate
MR times
aids
sig-
relatively of
diagnosis tumor
September
from
1990
a.
Ti-weighted large enhanced the ora serrata (open arrow)
b.
Figure 4. (550/i7)(b) the tumor
Transverse proton-density images show the plaque margin. Contrast between
(2,500/25) (a) and Gd-DTPA-enhanced Ti-weighted (arrow) eccentrically positioned with its edge abutting the plaque and tumor is comparable on both sequences.
ly,
(550/17) image showing a lesion extending anterior to (solid arrow). The plaque does not extend as far anterior-
resulting
in
incomplete
coverage.
References 1.
Packer
2.
3.
HD,
5.
6.
7.
effusion
8.
stays
bright
and
blends
with
vitreous
Iodine-125
Fitzpatrick
Peyman
fluid.
Peyster Hershy diology
9.
fluid and subretinal allowing more preof the tumor bound-
aries.
The
relationship
of the tumor
to the
plaque was readily appreciated on all 15 studies, despite the fact that two tumoms extended anterior to the oma serrata. Tumors as small as 3 mm thick
were also detected with MR imaging. Incomplete coverage of one lesion shown at MR imaging was due to maccurate preoperative measurement of the actual tumor base diameter with
maining
14 cases
obtained
intervals
showed
30%-40%
at
in
tumor volume. All patients had stabilization of their visual acuity, and none developed optic nerve atrophy, including the patient in whom the proximity of the tumor to the optic nerve resulted in plaque tilt and eccentric placement. In conclusion, MR imaging is a safe and effective method for assessing the relationship of the 1-125 plaque to choroidal melanoma and may be employed
when intraoperative able. MR imaging
US is not availbefore and soon after
US. This probably occurred because the tumor extended anterior to the ora serrata, a region not well depicted with US (4). The plaque, however, was not repositioned because of the clinical suggestion of scleral tumor extension,
plaque
may
which
MR in determining
later
Sequential
Volume
necessitated
enucleation.
US examinations
176
#{149} Number
of the
3
re-
when
placement
tumors
extend
PJ.
Ultrasound
determina-
GA, Mafee
MF.
Uveal
RG, BL,
10.
1 1
JJ. Shields
Augsburger Eagle R Jr.
Haskin
evaluation
with
1988;
melanoma
of magnetic tomography. 25:471-486.
JA,
ME.
MR
lntraocu-
imaging.
Ra-
168:773-779.
.
Lambrecht
L, Allewaert
R, de
Laey
braeken H, Brittoun J, van de Velde field resolution magnetic resonance of malignant choroidal melanoma. Ophthalmol
12.
reso-
Wilms G, Marshal G, Decrop E. van Hecke P. Dralands G. Surface coil magnetic resonance imaging of the orbit at 1.5 T. ROFO 1988; 149:496-501. Haik BG, Saint Louis L, Smith ME, Ellswirth RM, Deck M, Friendlander M. Magnetic resonance imaging in choroidal tumors. Ann Ophthalmol 1987; 19:218-222.
3-month
reduction
of posterior
and similar lesions: the role nance imaging and computed Radiol Clin North Am 1987;
lan tumors:
adjacent vitreous effusions (5-11), cise delineation
radiation
tion of the relationship of radioactive plaques to the base of choroidal melanomas. Ophthalmology 1989; 96:538-542. Williams DF, Mieler WF, Lewandowski M, Greenberg M. Echographic verification of radioactive plaque position in the treatment of melanomas. Arch Ophthalmol 1988; 106:1623-1624. Mafee MF. Peyman GA, Peace JH, Cohen SB, Mitchell MW. Magnetic resonance imaging in the evaluation and differentiation of uveal melanoma. Ophthalmology 1987; 94:341-348. Chambers RB, Davidorf FH, McAdoo JF, Chakeres DW. Magnetic resonance imaging of uveal melanomas. Arch Ophthalmol 1987; 105:917-921.
4.
Figure 5. Sagittal Ti-weighted (550/17) (a) and T2-weighted (2,500/90) (b) images. (a) On the Ti-weighted image, both the choroidal melanoma (solid arrow) and subretinal effusion (open arrow) are bright. (b) On the T2-weighted image, melanoma is dark while subretinal
S.
uveal melanoma. Ophthalmology 1987; 94:1621-1626. Bosworth JL, Packer 5, Rotman M, Ho T, Fingen PT. Choroidal melanoma: 1-125 plaque therapy. Radiology 1988; 169:249-251. Pavlin CJ, Japp B, Simpson ER, McGowan
Houdek
1988;
PV,
MR technique tion procedures IntJ RadiatOncol 1114.
JJ. VerE. High imaging
mt
11:199-205.
Schwade
JG,
Medina
for localization in episcleral Biol Phys
AJ, et al.
and
verifica-
brachytherapy. 1989; 17:1111-
be indicated
anterior
to the ora
serrata, which occurs in 13% of cases (6). A comparative study between MR imaging and US is still needed to more precisely define the clinical utility of tionships.
plaque/tumor
rela-
U
Radiology
#{149} 853