Cancer of unknown primary site

Abstract

Cancer of unknown primary site is one of the most challenging problems for physicians treating patients with cancers. Unfortunately we know relatively little about cancer of unknown primary site. We do not know which diagnostic tests and therapies are most effective. Much research should be done, but no controlled clinical trials are under way. The definition of a cancer of unknown primary site, or occult pr imary malignancy, requires the presence of a biopsy-proven malignancy together with negative results from a history, a physical examination, appropriate radiological tests, blood counts, and urine and stool analyses. Any abnormalities found during these diagnostic procedures must be followed up in a logical way. For example, if there is blood in the stool, a GI work-up is indicated. I f there is blood in the urine, a work-up of the urologic system is undertaken. If after a reasonable workup, the primary site is not established, one is dealing with a malignancy from an unknown primary site. Certain assumptions are often made regarding unknown primary malignancy. The first is that the ultimate proven primary sites in such patients occur in the same proport ions as cancers in the general population. That is, one would find lung, breast, colon, and prostate cancers as the most common types. This assumption is not correct. The data of the End Results Group for the distribution of certain cancers in the United States show that the proport ion of lung cancer cases among patients with an unknown primary site is twice as large as that in the general populat ion of American cancer patients. However, breast cancer, which occurs in the End Results Group in substantial numbers, appears infrequently as an unknown primary, accounting for only 2-3% of cases. Likewise, colorectal cancer is less common as an unknown primary, but liver cancer is 10 times more com-