In 1989, we reported the immunologic identification of a prognostic molecule in the tumor cells of breast cancer patients with a poor prognosis for recurrence and death due to their disease (1). This prognostic factor was statistically independent ofimportant clinical parameters including tumor size, lymph node involvement, and assessment of estrogen and progesterone receptors; subsequent studies also showed it to be independent of other prognostic molecules including c-erb-B2 and cathepsin D (2). Tumors marked by this prognostic molecule were nearly 4 times more likely to recur and metastasize than tumors not so marked, representing a prognostic power as strong as the presence of cancer in the axillary lymph nodes ofpatients with T1 or T2 primaries (1-3).We