INTRODUCTION
Between
May, 1987 and June 1992, 113 cases of soft-tissue sarcoma
in the extremity were operated. 62 patients had post-operative
MR studies performed and preliminary data and statistical
considerations of post-operative fluid collections noted within
this group is presented. Data analyzed included location of
tumor, type of surgery, pre- and post-operative adjuvant therapy
and enclosing length, width and depth measurements. The data
were analyzed for basic statistics and factors relating to
presence of fluid using SPSS (SPSS, Inc., Chicago, IL).
Of
the 62 patients in the study, 21 exhibited post-operative
fluid collections (18 seromas, 3 hematomas). Of the 37 with
lesions in the thigh, 19 (51.4%) had a fluid collection
while only 2 out of 17 with lesions of the leg (11.8%) had
post-operative fluid (P=0.0052). No fluid collections were
identified in the upper extremity cases (6 wide local excision,
2 excision). Of 36 patients with wide local excision, 17
(47.2%) had a fluid collection while only 3 out of 20 with
excision (15%) had fluid collections (P=0.015). Four out
of five patients with pre-treatment (80%) including both
chemotherapy and radiotherapy had post-operative fluid while
only 17 out of 57 (29.8%) with no, chemotherapy alone, or
radiotherapy alone had fluid collections (P=0.041) Of the
15 patients with no post operative therapy, only one (6.7%)
had fluid while 20 out of 47 (42.6%) of those with post-operative
therapy had fluid collections. (P=0.0086).
There
were no significant factors relating to the tumor volume
as measured by either enclosing rectangle
or ellipsoidal
approximate
either
as is or log transformed to stabilize the variance.
Post-operative
fluid collections following surgery in extremity soft-tissue
sarcomas are common in the thigh and following wide local
excisions (preferred surgical approach). On MR imaging post-operative
seromas vary in size, may be retained at the post-operative
site for extended periods, may increase in size and the
internal characteristics of the fluid collection as noted
on MR imaging may vary over time. These MR imaging appearances
should be recognized to avoid confusion with recurrence
or other lesions.
Dennis
A. Johnston Ph.D. (Dept. of Biomathematics,
The University of Texas M. D. Anderson Cancer Center) provided
assistance with statistical analysis and Raphael E. Pollock,
M.D., Ph.D. (Dept. of Surgery, The University of Texas M.
D. Anderson Cancer Center) provided assistance with surgical
data collection. Their contributions are gratefully acknowledged.
