Downloaded from www.ajronline.org by 220.174.236.95 on 11/05/15 from IP address 220.174.236.95. Copyright ARRS. For personal use only; all rights reserved
85
Diagnostic Evaluation of Cancer Patients with Pelvic Pain: Comparison of Scintigraphy, and MR Imaging
.1.;‘; .:
Thomas E. Beatrous1 Peter L. Choyke2 Joseph A. Frank2
Pelvic pain in cancer patients can result choice of imaging techniques for evaluating evaluated
27
cancer
patients
(24 patients),
abdominal the correlation between
with
pelvic
CT (27 patients),
CT,
,
..
:.
.#{149}.
..
from several causes. The most appropriate such patients has not been established. We pain
by using
radionuclide
bone
scintigraphy
and pelvic MR imaging
(27 patients) and used symptoms and imaging findings to compare these imaging methods. The study population included I 1 patients with Ewing sarcoma, six with other sarcomas, five with colorectal cancers, and five with other tumors. All patients had pelvic pain, and eight had pain radiating to a leg. Twenty-three patients had soft-tissue masses, and 19 had bone metastases; 16 had both. Findings on bone scans explained the symptoms in 17 (71%) of 24 patients, findings on CT in 23 (85%) of 27 patients, and findings on MR imaging in 25 (93%) of 27 patients. The difference between bone scanning and CT or MR was statistically significant (p < .05); however, the difference between CT and MR imaging was not significant (p > .05). MR imaging detected 41 (98%) of 42 relevant lesions, whereas CT detected 31 (74%) of 42, and bone scanning 17 (44%) of 39. We conclude that MR is superior to either bone scanning or CT in the initial evaluation of pelvic pain in cancer patients. Such information can be important in directing the treatment of these patients. AJR
155:85-88,
Specific
July
causes
1990
of pelvic
pain
in cancer
patients
include
bone
and
soft-tissue
metastases or complications after surgery or radiation therapy, as well as benign disease such as disk hemiation. A thorough history and physical examination often can differentiate these causes, but imaging studies frequently are performed to define or exclude an anatomic lesion. The methods available include plain radiography, scintigraphy, CT, and MR imaging. We compared the sensitivity of three imaging procedures in identifying the causes of pelvic pain in the cancer patient. Received revision
November
February
This work
20, 1989; accepted
after
12, 1990.
was performed
in part
in the In vivo
NMR Research Center at the National Institutes of Health. 1
Radiation
Oncology
Branch,
National Cancer
Institute, National Institutes of Health, Bethesda, MD 20892. 2 Department of Diagnostic Radiology, Warren G. Magnuson
Center, National Institutes of MD 20892, and Department of
Clinical
Health, Bethesda, Radiology, Georgetown University Medical Center, Washington, DC 20007. Address reprint requests
to
P. L. Choyke, Diagnostic Radiology Department, Bldg. 10, Rm. 1C660, National Institutes of Health, Bethesda, MD 20892.
0361-803X/90/1551-0085 C American
Roentgen
Ray Society
Materials
and Methods
Twenty-seven patients with a known malignant tumor or myeloproliferative disorder were seen between April 1985 and January 1989 for evaluation of pelvic, sacral, or buttock pain. Eight also had sciatic pain. All 27 patients underwent CT and MR imaging, and 24 underwent bone scanning. Only patients who had two or more studies were included in this study. The procedures
were
performed
within
2 months of one another, most (23 of 27) within 2 weeks.
The order of the studies was random. There were 14 males (aged 14-65) and 13 females (aged 10-61). Underlying diagnoses in the 27 patients included Ewing sarcoma in 1 1 , other sarcomas in
six (osteosarcoma, lymphoma
chrondrosarcoma,
and rhabdomyosarcoma),
colorectal cancer in five,
in three, and other tumors in two (ovarian, melanoma). The pain was right-sided in 1 2 patients, left-sided in 1 1 , and bilateral in four. Sixteen patients had both soft-tissue masses and bone disease, seven had soft-tissue masses only, and four had bone disease only (two of these were from benign causes: bone cyst with fracture and avulsion fracture).
BEATROUS
86
Bone scans were obtained MDP by using
planar
with 1 5-20 mCi (555-740
imaging.
CT
was
performed
Downloaded from www.ajronline.org by 220.174.236.95 on 11/05/15 from IP address 220.174.236.95. Copyright ARRS. For personal use only; all rights reserved
on a 0.5-T Picker(Highland
Heights,
AJR:155,
July 1990
MBq) Tc-
GE 9800 unit
on a
(General Electric, Milwaukee, WI); 10-mm contiguous used. IV contrast material was used in 13 patients. formed
ET AL.
sections were MR was per-
OH) MR unit (23 patients)
and a 1 .5-T General Electric unit (four patients) with the following nominal pulse sequences: Ti -weighted spin echo, 300-500/12-20 (TR/TE),
in axial, sagittal,
2000/80-1
00, in the axial plane; and short inversion-time inversion (STIR), 1500-1 640/1 00-1 40/26-30 (TR/TI/TE), in the cor-
recovery
onal
plane.
Additional
and coronal
imaging
planes;
planes
T2-weighted
and
pulse
spin-echo,
sequences
were
used as needed. One radiologist performed a retrospective review of the bone scans, CT scans, and MR images without knowledge of the clinical symptoms. Lesions were identified on the basis of their anatomic proximity or involvement of the sacral nerve roots and sciatic nerve. Lesions not near the appropriate site of pain were not considered
relevant to this evaluation. Medical records were reviewed results.
Pain distribution
confirmatory
test
myography
(four
without
was recorded,
such
as
knowledge
myelography
patients).
The
of the imaging
as well as the results
results
(three
patients)
were
compared
of any
or electro-
then
with
imaging findings. When a lesion corresponded exactly or closely to the patient’s symptoms, the study was considered positive. If the study findings were normal or did not explain the symptoms, the study was considered negative. Statistical differences were deter-
A- Lumbar roots
mined
U. Sacral roots
by using
the McNemar
test.
. 0-
Results An anatomic
abnormality
that could
account
later), imaging findings were normal. MR imaging and CT were performed in 27 patients, and bone scintigraphy was performed in 24. Findings on MR imaging explained the symptoms in 25 (93%) of 27 patients, including 23 (100%) of 23 soft-tissue lesions and 18 (95%) of 1 9 bone lesions. Soft-tissue findings on MR imaging included cases
of cauda
equina
and nerve-root
metastases,
17
cases of neoplasm involving the lumbosacral nerves in the pelvis, two cases with involvement of the sciatic nerve in the buttocks, and two cases with involvement of more than one site (Fig. 1). Findings on CT explained the symptoms in 23 patients (85%), including 17 (74%) with 23 soft-tissue lesions and 14 (74%) with 1 9 bone lesions. Findings on bone scintigraphy
explained
the symptoms
showing no soft-tissue lesions. Of the 42 possible lesions
and 1 9 bone
in 1 6 (67%)
masses sites
lesions),
of 24 patients,
and 1 7 (89%)
of abnormality MR imaging
of 19 bone
(23 soft-tissue detected
and CT detected 31 (74%). Bone scintigraphy 1 7 (44%) of 39 possible sites.
41 (98%)
detected
only
No correlation was found between imaging findings and pain in two patients. One of these was a 1 4-year-old boy with
a history of Ewing sarcoma of the left humerus who had pain in the right buttock. Radiographs showed an abnormality in the
right
showed findings initially
ischial
tuberosity
suggestive
of
a fracture that was misinterpreted on bone scans were nonspecific. were
interpreted
vealed an avulsion
as normal.
malignancy.
CT
as a tumor. The Findings on MR
Surgical
injury of the ischial tuberosity.
Sciahc notch cuteal
for the pa-
tients’ symptoms was present on at least one of the imaging techniques in 26 of 27 patients. In one patient (discussed
two
-
resection
re-
The second
Fig. 1.-Sites
of pelvic-nerve
metastases
(n
=
23).
patient had ovarian cancer with left-sided sciatic pain, but findings on MR images, CT scans, and bone scans were normal. One and one-half years after MR imaging, the patient was cancer-free; there was still no explanation for her symptoms, which had improved. In 16 of 21 patients in whom nerve-root involvement was present, the coronal STIR images provided superior images of the anatomy compared with spin-echo Ti - and T2weighted images and CT. All cases of clinically suspected nerve-root involvement were confirmed on STIR images, although pathologic correlation was not available. Discussion Low-back pain or sciatica can result from several sources. In addition to benign causes such as disk hemiation, the oncologic patient is at risk for the development of a tumor or metastasis that directly affects the nerve roots and sacral plexus. The vague or confusing nature of symptoms can make the clinical diagnosis of an anatomic abnormality difficult [1]. Moreover, the intricate anatomy of the lumbosacral plexus can complicate diagnostic imaging [2, 3]. Imaging techniques such as bone scanning and CT are useful in the evaluation of patients with sciatic symptoms [4]. However, scintigraphy is limited to processes affecting the skeleton, and CT can fail to show retropentoneal tissue masses
that
over, subtle
are too
small
(Figs. 2-4)
to be considered
soft-tissue
pathologic.
or bone lesions
More-
may be
PELVIC
Downloaded from www.ajronline.org by 220.174.236.95 on 11/05/15 from IP address 220.174.236.95. Copyright ARRS. For personal use only; all rights reserved
AJR:155, July 1990
Fig. 2.-Right
lumbosacral
metastases
PAIN
IN CANCER
PATIENTS
87
In-
volving right lumbosacral plexus in patient with melanoma. A, Bone scintigram shows focus of abnormal activity (arrow) in right L4 vertebral body. Sacrum was considered normal. B and C, Abdominal CT scans show normal sacrum and only mild enlargement of right piriformis muscle (arrow). D, Coronal STiR MR image (1500/100/26) shows metastatic lesion at L4 (straight arrow) and sacrum (curved arrow). E, More posterior image shows involvement of lumbar roots (straight solid arrow) and sacral nerve roots (curved arrow). Contralateral normal sacral nerve root (open arrow) is well seen.
A Fig. 3.-54-year-old
patient with metastatic
colorectal cancer and metastases
involving lumbosacral
A, Bone scintigram shows no abnormalities. B, Abdominal CT scan shows two soft-tissue
as unopacified
C, Coronal STiR MR image (1500/100/26)
arrows)
(M) is seen also in rectosigmoid
colon.
masses (arrows) initially misinterpreted shows multiple soft-tissue masses (straight
roots. bowel.
in course
of sciatic
nerve (curved
arrows).
Primary mass
BEATROUS
Downloaded from www.ajronline.org by 220.174.236.95 on 11/05/15 from IP address 220.174.236.95. Copyright ARRS. For personal use only; all rights reserved
88
ET AL.
AJR:155, July 1990
malignant tumors. Pelvic MR alone centage of lesions when compared ning in the diagnosis of lumbosacral Nevertheless, we wish to emphasize study and the population from which sent those with the most advanced thus
Fig. 4.-20-year-old patient with Ewing sarcoma of pelvis and lower back pain. STiR MR image (2000/160/25) at 1.5 T shows discrete metestases to right L4 and L5 nerve roots (black arrows). Note normal-appearing roots on contralateral side (white arrows). Findings on CT and bone scintigraphy were normal in this area.
were
more
on CT. MR is useful
conditions in the
such
early
as disk
detection
herniation
of vertebral
in the diagnosis and lumbar metastases,
of benign stenosis
because
-
on the STIR sequence to depict normal and abnormal neural and osseous anatomy, and this pulse sequence was crucial to our success with MR images. The STIR sequence, first described by Bydder and Young [6], uses an inverting pulse effectively
eliminates
the contribution
of fat to the MR
image. Thus, bone marrow and retroperitoneal fat are low in signal intensity on STIR images [7, 8]. Areas of malignancy, in general,
are high in signal
and conspicuous
against
the fat-
suppressed background [9]. Moreover, the STIR sequence provides excellent delineation of the proximal nerve roots and cord (Figs. 2-4). The sciatic nerve itself may not be seen, but its location can be inferred from signal arising within the accompanying inferior gluteal artery and vein. Parenthetically, field
useful
strength
scans
is not an inherent
were
obtained
limitation,
at both
positive
findings
than
would
ACKNOWLEDGMENTS
the manuscript,
We thank Anna Scheib for preparing assistance,
and Eli Glatstein
for reviewing
Cathy Chow the manu-
script.
and of its
multiplanar capabilities and ability to show disease processes [5]. Standard Ti and T2-weighted spin-echo images, however, do not show nerve roots well. In this study, we relied
that
to yield
patients from a more general oncologic setting. Also, when MR imaging is unavailable or contraindicated, CT and bone scintigraphy usually will suffice. MR combines in one study the ability to assess disk herniation, metastasis to the bone, and nerve-root involvement, all of which can cause pelvic pain. This provides a rationale for the use of pelvic MR as the primary means of evaluating pelvic pain in oncologic patients. The findings of these studies also can be used as a guide for appropriate radiation therapy or follow-up of chemotherapy.
for technical
missed
likely
showed a greater perwith CT and bone scannerve-root involvement. that the patients in this they were drawn represtages of disease, and
as diagnostically
0.5 and 1 .5 T in this
study.
Metastases to the pelvis and nerve roots are a frequent cause of pelvic pain, particularly in patients with advanced
REFERENCES 1 . Scher
HI, Nagoda
A. Bone metastases:
for use of resorption inhibitors.
rationale 6-28
2. Gebarski
KS, Gebarski
lumbosacral RadioGraphics
plexus:
SS, Glazer GM, Samuels
anatomic-radiologic-pathologic
1986;6(3):401-425 3, Pech P, Haughton v. A correlative nerve. MR 1985;144:1037-1041 4. Cohen BA, Lanzien CF, Mendelson sciatic
foramen
pathogenesis, treatment 1987;82[Suppl
Am J Med
in patients
BI, Francis correlation
CT and anatomic
and
2Aj:
IR. The using
CT.
study of the sciatic
DS, et al. CT evaluation of the greater
with sciatica.
AJNR
1986;7:337-341
5. Stack J, Redmond 0, Scully M, Camey D, Ennis JT. Assessment of bone marrow involvement in small cell lung cancer with spin echo and gradient echo sequences. Presented at the annual meeting of the SOCiety of Magnetic Resonance in Medicine, San Francisco, August 1988 6. Bydder GH, Young IR. MR imaging: clinical use of the inversion recovery sequence.
J Comput
Assist
Tomogr
1985;9:659-675
7. Dwyer AJ, Frank JA, Sank vJ, Reinig JW, Hickey AM, Doppman JL. Short TI inversion-recovery pulse sequence: analysis and initial experience in cancer
imaging.
Radiology
1988;168:827-836
8. Shields AF, Porter BA, Churchley 5, et al. The detection of bone marrow involvement by lymphoma using magnetic resonance imaging. J Clin Oncol 1987;5:225-230
9. Frank JA, Dwyer AJ, Doppman JL. Magnetic resonance imaging in oncology. In: Devita ‘/T, Hellman 5, Rosenberg SA, eds. Importantadvances in oncology 1987. Philadelphia: Lippincott, 1987: 133-1 74