Ten endometrial stromal tumors of the uterus with a prominent myxoid or fib
rous appearance, or both, that led to problems in interpretation are report
ed. The patients were 32 to 52 (mean 39) years of age. Three presented with
dysfunctional uterine bleeding and one with abdominal pain. An enlarged ut
erus or a pelvic mass was palpated in five patients; the tumor was an incid
ental postpartum finding in one patient. All patients underwent hysterectom
y. The tumors ranged from 4 to 20 cm in greatest dimension. Six were soft.
polypoid intracavitary masses and four were predominantly intramyometrial:
two were gelatinous. On microscopic examination, nine tumors infiltrated th
e myometrium (stromal sarcomas) and one was well circumscribed (stromal nod
ule), Six tumors had a predominantly fibrous component with the neoplastic
cells separated by variable amounts of collagen: extensive areas of hyalini
zation were present in three tumors. Two tumors were predominantly composed
of hypocellular areas with an abundant myxoid matrix, and two had both com
ponents in roughly equal proportions. Alcian blue staining was positive, wi
th the staining eliminated by hyaluronidase predigestion. in the myxoid are
as. The typical morphologic features of endometrial stromal neoplasia were
present focally in four tumors. All of them contained numerous small thin-w
alled vessels. Vimentin and smooth muscle actin were positive in nine of ni
ne and seven of nine tumors, respectively, whereas desmin was negative in s
ix of nine tumors and only focally positive in the other three. One patient
had omental nodules at the time of the initial diagnosis and another had a
pelvic recurrence 2 years after hysterectomy. Follow-up information is una
vailable or short in the other cases. These tumors should be considered of
endometrial stromal origin in view of the typical location of most of them,
their growth pattern, content of characteristic arterioles. presence of ty
pical endometrial stromal neoplasia in the primary or recurrent tumor in so
me cases, and absence of evidence of origin from a cell type other than end
ometrial stroma. These rumors may be identical, in some instances at least,
to tumors referred to in the older literature as "myxofibrosarcomas."