Breast cancer as analyzed by the human tumor stem cell assay.

Patients with primary and recurrent carcinomas of the breast were studied by the human tumor stem cell assay to determine if (1) colonies would form from breast cancer specimens, (2) growth in the culture would equate with aggressiveness of disease, (3) the assay would yield specific information on drug responsiveness, and (4) the assay would yield nonspecific information on drug responsiveness. Colony counts ranged from 0 to 363. There was no significant difference in median colony counts by pathologic stage of disease or site. Among stage IV patients presenting for treatment with primary disease, those with colony counts greater than 10 had a mortality rate of 4.7/1000 person-days; there were no deaths among those with colony counts less than or equal to 10 (P = 0.042). Stage IV patients presenting with recurrent disease showed no association between colony counts and survival (P = 0.53). No significant relationship between colony counts and disease-free intervals was observed among stages I, II, and III patients (P = 0.10). Drug sensitivity in vitro was found in 14% of the cultures with colony counts greater than or equal to 30. The only complete clinical responses in stage IV patients occurred in two patients with 0 colony counts. These data demonstrate that colonies grow from breast cancer specimens, that colony formation in vitro may be related to aggressiveness of growth in vivo in patients presenting with stage IV disease, that drug sensitivity is demonstrated in few cultures, and that patients with metastatic disease who have complete response to systemic therapy may be identified by lack of growth in the culture.