|
Case Presentation
Carcinoma of
Unknown Primary
By Janet Robbins Hosenpud, M.D.; Assistant Professor of
Medicine
G.A. is a 28-year-old Hispanic man who presented to a health care
provider in the summer of 1994 with complaints of chest tightness
unrelated to exertion or breathing. A CXR revealed hilar and mediastinal
adenopathy, and a CT scan essentially confirmed these findings, plus
demonstrated a paraspinal mass in the T3 to T5 region. Mediastinoscopy
showed reactive hyperplasia of the nodes, as well as features of sarcoid.
He was started on Prednisone, which was slowly tapered during the next
three months. Over the next several months he noted slowly progressive
mid-thoracic back pain. Finally, in November of 1994 he presented to his
health care provider with lower extremity parasthesia and rapidly
progressing lower extremity weakness. He was transferred to Froedtert
Hospital for further care, and was admitted to the neurosurgery
service.
On admission, he was noted to be completely paraplegic, unable
to move or feel from the level of umbilicus. An MR of the spine revealed a
thoracic paraspinal mass and evidence of cord compression. An angiogram
revealed the mass to be hypervascular. A fine needle aspirate of the mass
showed clusters of small round cells consistant with a "neoplasm." The
mass was angiographically embolized and 24 hours later the patient was
taken to the operating room where he underwent excision of the posterior
mediastinal mass, partial T3 and T4 corpectomies, anterior spinal cord
decompression, a T2 to T3 anterior fusion, and a T2 to T10 posterior
fixation. The pathology of the mass was consistant with a epithelial
malignancy. PAS stains showed little or no glycogen in the tumor cells.
Immunoperoxidase stains revealed: leukocyte common antigen negative,
chromografin negative, low molecular weight cytokeritin positive, factor
VIII negative, MMB 45 negative, neurofilament negative. Electron
microscopy showed the cells to be poorly differentiated with rare
junctional complexes, there were no glycogen aggregates or lipid-glycogen
aggregates typical of Ewings sarcoma, no cytoplasmic inclusions consistant
with dense-core neurosecretory granules were seen.
Between December
1994 and February 1995 the patient received 4 cycles of chemotherapy
consisting of Ifosfamide, cis-Platinum, and Etoposide. During this time he
was followed closely be the rehabilitation service, and experienced
progressive improvement in his lower extremity strength, and progressive
normalization of sensation. He went from being wheelchair-bound, to
walking into the clinic with his children. CT scans of the thoracic spine
have shown no abnormalities other than post-operative changes, which have
been stable.
It is estimated that carcinoma of unknown primary occurs
in 3 to 15% of all patients referred to medical oncologists. Despite
advances in immunocytochemistry and the use of monoclonal antibodies to
help with our assessment, these neoplasms continue to bedevil oncologists.
In fact, the "unknown primary carcinoma" (UPC) is a distinct entity in the
medical oncology literature. While traditional teaching in medical schools
is that an exhaustive search for a primary site in these patients is not
warranted, a recent review of these cases and their treatment argues that
limited searches may benefit some patients. (1) This review 879 patients
referred to MD Anderson revealed that of the patients whose primary site
COULD be found, (20% of the total), there was improved survival (11 months
vs. 15 months). The group with known primary tumors had a 10% better
survival each year. The improvement in outcome of the "known" UPCs was
mainly the result of finding treatable cancers such as lymphoma, breast
cancer, ovarian cancer and germ cell malignancies. Almost a third of the
"unknown primary cancers" were able to be identified simply by a complete
pathology review. The identification of a group of unsuspected lymphomas
was a major reason why these patients had a relatively better
outcome.
In a previous review by the same group (2), a sub-set of
patients with a favorable prognosis were confirmed: those with axillary
node involvement, and those with neuroendocrine cancer. Patients with
poorer outcomes were males with non-neuroendocrine hepatic metastasis, and
supraclavicular lymph node metastasis. Not suprisingly, it appeared that
patients presenting with multiple metastatic lesions had a poorer outcome.
These reports seem to confirm a survival advantage to patients with the
gamut of more treatable cancers such as breast cancer, small cell
carcinoma, lymphoma and ovarian cancer compared to the patients with the
list of less responsive cancers such as pancreas, lung, or melanoma. Using
the CT scan to definitively find the pancreatic cancer does not make it
more responsive to treatment.
The younger, non-smoking patient with
predominantly retroperitoneal and/or "midline" tumor involvement is in yet
another group that are felt to have a particularly good outcome. In a
review of UPC patients that seemed to be heavily weighted toward this
population, Hainsworthe et all found a 63% response to cis-platinum based
chemotherapy, and actuarial survival at 10 years was a relatively
impressive 16%. (3) It appears, then, that a certain number of these
patients will respond to and definitely benefit from both a recognition of
their disease and systemic treatment. The guidelines in these patients
seems to be first, do the simple tests that will simply find the most
treatable malignancies (mammogram, chest x-ray, a complete physical exam
including pelvic and rectal exams). Second, the pathology review is
indispensable. Third, watch for the young man with predominantly midline
disease, he has a chance for a long life.
|
