The Significance of Detecting Cytokeratin-Positive Cells in Breast Cancer
ABSTRACT & COMMENTARY
Source: Braun S, et al. N Engl J Med 2000;342:525-533.
The presence of cytokeratin-positive cells in the bone marrow in patients with breast cancer may be a sensitive indicator of occult metastases. Braun and colleagues from Munich have recently reported results from the analysis of bone marrow aspirates from 552 newly diagnosed patients with stage I, II, or III breast cancer and 191 control patients with non malignant conditions. Of the patients, 199 (36%) had cytokeratin-positive cells in the bone marrow at the time of the initial resection of the primary tumor. Cells that stain positively for cytokeratin are considered of epithelial origin and are not usually found among cells aspirated from the bone marrow. In a patient with cancer, such cells are likely to represent occult metastases.
In this series, the overall frequency of occult metastatic cells in each specimen was low; there was a median of three cytokeratin-positive cells (range, 1-1223) per 2 ´ 106 bone marrow cells analyzed. The numbers of detectable tumor cells increased with the tumor stage. For example, patients with stage I cancer had a mean of five tumor cells per 2 ´ 106 bone marrow cells, whereas patients with stage II and III disease had means of nine and 86 tumor cells per 2 ´ 106 bone marrow cells, respectively.
Bone marrow aspirates from 191 patients with nonmalignant disease were also analyzed and only two—one with a chronic benign inflammation of the breast and the other with a benign cystadenoma of the ovary—were cytokeratin-positive.
In the breast cancer series (552 newly diagnosed patients), the majority (58%) had small (< 2 cm) tumors. However, larger primary tumors were associated with a higher incidence of micrometastases than were tumors that were 2 cm or less in diameter (P < 0.001). Of the 19 patients with stage T4 tumors and inflammatory characteristics, 15 had occult metastatic cells in the marrow (P < 0.001).
Surprisingly, there was no statistical correlation with the presence or absence of lymph node involvement. Of the 301 patients without clinical or histological evidence of node involvement, 100 (33%) were marrow cytokeratin cell positive. Nonetheless, for those with nodal metastases, there was a direct correlation of the number of nodes positive and the presence of cytokeratin-positive cells in the marrow.
At the time of this report, 552 patients had been followed for a median of 38 months (range, 10-70) and relapse had occurred in 135 (28 with locoregional and 107 with distant metastases). Whereas locoregional relapses were not associated with the presence of micrometastases in bone marrow (at the time of initial diagnosis), distant metastasis was significantly associated with the presence of occult micrometastases in the marrow. Of 33 patients with relapses at visceral sites alone (without skeletal metastases), 13 had bone marrow involvement initially. In contrast, in 18 of 19 patients with relapses in the skeleton and in 48 of 55 patients with relapses at visceral sites in combination with skeletal metastases, micrometastases had been detected on the initial bone marrow (P < 0.001).
The presence of cytokeratin-positive cells in the marrow did confer some prognostic information. Of 199 patients with occult metastatic cells, 49 died of cancer-related causes (25%), whereas of 353 patients without occult tumor cells in the marrow only 22 died of breast cancer (6%). Patients with bone marrow micrometastases had a higher risk of death from cancer than patients without bone marrow micrometastases (relative risk, 4.28; 95% confidence interval, 2.59-7.09; P < 0.001). Among women with cytokeratin-positive cells in the marrow, as compared to those without such cells, the relative risk of death was 3.32 among patients with node-positive cancer (95% confidence interval, 1.91-5.76; P < 0.001) and 13.26 among patients with node-negative cancer (95% confidence interval, 3.01-58.46; P < 0.001). Among 100 patients with node-negative cancer and micrometastases, 14 (14%) died of cancer-related causes, whereas only two patients (1%) died of cancer-related causes in the group of 201 node-negative patients without micrometastases.
COMMENT BY WILLIAM B. ERSHLER, MD
This is an important paper for a number of reasons. First, it reports a fairly high incidence of systemic (bone marrow) micrometastases in newly diagnosed, node-negative breast cancer patients (33%). Second, it shows that those with bone marrow micrometastases are more likely to relapse than those without. Furthermore, their survival is clearly shorter when compared to similarly staged patients but without marrow cytokeratin cells.
Is this useful information? It may well be. Bone marrow analysis for cytokeratin-positive cells might allow a more educated prediction of relapse risk, and may therefore help determine those who might benefit from adjuvant chemotherapy. Node-negative patients with a small tumor might not typically be treated with adjuvant chemotherapy. However, with the knowledge that the marrow contains occult metastases, institution of systemic therapy would seem justified.
The presence of cytokeratin-positive cells in the marrow might also be a useful discriminator for those who might benefit from bisphosphonate infusions. Considerable enthusiasm has developed for these agents in treatment of patients with metastatic disease to bone,1 and it has been suggested that early treatment may delay or even prevent clinically important bone disease.2
The downside of this approach is that bilateral iliac crest marrow aspirates are not routinely performed in the initial evaluation of breast cancer, and its introduction will be the cause of some morbidity and additional expense. The data from Braun et al are intriguing but the clinical applicability, although intuitive, has yet to be established. Accordingly, staging marrows to detect cytokeratin-positive cells should remain investigational for the time being. Additional studies are needed to further define the value of bone marrow cytokeratin staining in the breast cancer staging. Eventually, it is possible that this information will have as much prognostic value as the examination of axillary nodes.
References
1. Hillner BE, et al. J Clin Oncol 2000;18:1378-1391.
2. Lipton A, et al. Cancer 2000;88:1082-1090.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content