dcyphr | Cytokeratin-positive cells in the bone marrow and survival of patients with stage I, II, or III breast cancer.


Braun et al. tested the effect of cytokeratin-positive cancer cells on the development of breast cancer. The authors extracted bone marrow aspirates from the upper iliac crests of patients, with stage I, II, or III breast cancer, who had the tumor removed. Bone marrow aspirates were also extracted from non-cancerous controls. The aspirates were stained with a monoclonal antibody, A45-B/B3, which is bound to antigens on cytokeratins. Follow up appointments ranged from 10-70 months and the end goal was survival. Cytokeratin-positive cells were found in the bone marrow of one percent of the controls and thirty-six percent of the patients with breast cancer. Braun et al. found that the presence of occult metastatic cells in bone marrow was unrelated to that of lymph-node metastasis. 49 patients out of the 199 patients with occult metastatic cells died of cancer, whereas out of the 353 patients without such cells, 22 died of cancer-related causes. Out of 301 women without lymph-node metastases, 100 had bone marrow micrometastases. 14 of the 100 died of cancer-related causes, as did 2 of the 201 without bone marrow micrometastases.  These results show that the presence of occult cytokeratin-positive metastatic cells in bone marrow increases the risk of relapse in patients with breast cancer.


It is important to track the early spreading of cancer cells because this is one of the biggest causes of relapse and death. These cells influence the spread of cancer in patients. The presence of ectopic epithelial cells in the bone marrow was shown to influence the development of patients with colorectal, gastric, and non–small-cell lung carcinomas, tumors. In breast cancer, antibodies against antigens of the polymorphic epithelial mucin family have been used to identify ectopic epithelial cells. Results from a study of 49 patients showed that cytokeratin-specific antibodies were able to detect breast-cancer micrometastases from bone marrow smears.  This prompted Braun et al.  to study a defined number of bone marrow cells, and a monoclonal antibody against an antigen shared by various cytokeratin peptides. The results of Braun et al.’s study demonstrates that cytokeratin-positive cells in bone marrow are indeed tumor cells.


From January 1994 to December 1997, the I. Frauenklinik at Ludwig Maximilians University in Munich and the Zentralklinikum in Augsburg, Germany, studied 743 patients. These patients had bone marrow aspirates taken from both upper iliac crests before removing the cancer. Braun et al. used this method to examine the bone marrow obtained from 552 patients with stage I, II, or III breast cancer. Of the 552 patients with breast cancer, 298 patients went through breast conservation 254, modified radical mastectomy.  All patients had the tumor removed. Out of the patients treated with breast-conserving surgery, 298 received radiation therapy. 72, of the 170 postmenopausal women with node-positive breast cancer (with estrogen-receptor–positive tumors) received 20 to 30 mg of tamoxifen daily. Both premenopausal and postmenopausal patients with estrogen-receptor–negative tumors were treated with chemotherapy.

The base-line diagnostic evaluation for distant metastases, at the time of primary surgery, included plain chest radiography, mammography of the contralateral breast, ultrasonography of the liver, and bone scanning of the entire body. Patients were met with clinical examinations every three months after surgery and were only tested further if they showed any symptoms.

Braun et al. obtained bone marrow samples from the upper iliac crests of the patients by needle aspiration. during primary surgery and stored in heparin-treated tubes.

The authors screened 2×106 cells by bright-field microscopy. All slides they produced were examined separately by two different observers who agreed on the results for over 95 percent of specimens.

Braun et al. used monoclonal antibody A45-B/B3 to find tumor cells in cytospin preparations of bone marrow. The authors used the breast-cancer cell line BT-20 as a positive control for cytokeratin immunostaining and they used the log-rank test to compare the patients with bone marrow micrometastases against those without micrometastases. Braun et al. used Cox proportional-hazards analysis to estimate the effect of various variables on the development of the disease, and they used the chi-square test to compare categorical variables. Finally, for statistical analyses, the authors used SPSS software for Macintosh.


199 of the 552 patients, who had bone marrow aspirates taken, had cytokeratin-positive tumor cells in the bone marrow at the time of the initial resection of the primary tumor. In the majority of the specimens, occult cells were present as dispersed single cells and the frequency of occult metastatic cells was low. Bone marrow aspirates from the 191 controls were also analyzed before the final result was disclosed. Stained cytokeratin-positive cells were detected in only two of the 191 patients. 

58 percent of the patients had primary tumors that were no more than 2 cm in diameter. Primary tumors that were larger than 2cm were more likely to have a higher count of micrometastases. Of the 43 patients with stage pT4 tumors, 19 had inflammatory breast cancer; of the 19, 15 had occult metastatic cells in the bone marrow. Twenty-three percent of patients with stage pT1a tumors, 35 percent of patients with stage pT1b tumor, and 30 percent of patients with pT1c tumors had occult disease. 

Braun et al. found that the incidence of bone marrow micrometastases was similar in patients with lymph-node metastasis and those without it. 


After a 10 to 70 month range of follow-ups, 135 patients experienced a relapse of the tumor. 28 of the 135 patients had a locoregional relapse, and 107 had distant metastases. Locoregional relapses were not associated with the presence of micrometastases in bone marrow; however, distant metastases were.


49 patients with occult metastatic cells, out of 199, died of cancer-related causes. On the other hand, out of 353 patients without occult tumor cells in the marrow, only 22 died of breast cancer. 14 patients died of cancer-related causes out of 100 patients with node-negative cancer and micrometastases. However, only 2 patients died of cancer-related causes out of the 201 patients without micrometastases. Braun et al. found no significant difference in survival between patients with node-negative cancer who had micrometastases and patients with node-positive cancer who did not have micrometastases.


Braun et al. analyzed 245 patients with node-negative cancer who didn’t receive adjuvant therapy separately due to the possibility of locoregional relapse and distant metastasis being  influenced by adjuvant treatment. 81 of the 245 had occult metastatic cells. Of the patients, who didn’t receive Adjuvant therapy, the risk of cancer-related deaths was higher of the 81 patients with micrometastases rather than of the 164 without micrometastases.


The authors found that bone marrow micrometastasis, estrogen receptors, and lymph-node metastasis were each independent predictors of both recurrence with distant metastases and cancer-related death through a Cox multiple-regression analysis.


Braun et al.’findings support the view that different pathways of tumor-cell dissemination cause distinct patterns of metastasis this is supported by the studies which found no concordance between the presence of lymph-node metastasis and the presence of bone marrow micrometastases.