Circulating Tumor Cells and Biomarkers: Implications for Personalized Targeted Treatments for Metastatic Breast Cancer

Abstract

To the Editor:Metastatic breast cancer (MBC) is incurable andthe median survival of affected patients ranges from 2to 4 years with a few patients surviving for more than5 years (1–3). Palliation of metastatic disease can beachieved by considering pretreatment prognostic andpredictive factors, which can assist in choosing a ther-apy with the greatest likelihood of patient benefit (1).Another factor that may accurately predict prognosisand treatment efficacy in patients with MBC is thenumber of peripheral blood circulating tumor cells(CTC) (4), thereby providing a patient risk stratifica-tion tool.Recent studies confirmed the superiority of CTCover imaging assessment and other standard measures,such as hormone receptor status and tumor burden,for independently predicting survival (5). Thus, CTCmay indicate the existence of cancer cells that have aunique phenotype that distinguishes them from thepredominant population of cancer cells and raises thepossibility of resembling cancer-initiating stem cells.Hence, the characterization of CTC might provideinsights into the mechanisms of tumor developmentand metastases, and identify potential novel targetsfor therapy.In breast cancer, CD44(+)CD24()⁄low)Lineage())cancer-initiating cells are known to be highly tumori-genic in severe combined immunodeficient mice (6).Previous studies have shown that micrometastaticdisease detected in the bone marrow of patients withprimary breast cancer have CD44(+)CD24()⁄low)phenotype suggestive of putative cancer stem cells (7).The extension of these evaluations to cancer-initiatingcells is feasible and quite appealing for the possibletherapeutic implications. Gene expression profiles ofCTC would provide an opportunity for specific andindividualized treatment planning.We evaluated CTC from 20 women enrolled in aprospective clinical trial at The University of TexasM. D. Anderson Cancer Center for transcripts ofestrogen receptor (ER), progesterone receptor (PgR),and human epidermal growth factor receptor-2 (HER-2). Such biomarkers’ expression in CTC is currentlynot considered in the selection of treatment regimens,and hence we compared the expression patterns ofCTC with the expression of these proteins by the pri-mary tumor and metastatic lesions. Tumor character-istics, types of treatment, and responses for the 20women with MBC who contributed samples for thisanalysis are shown in Table 1. The median number ofCTC was 2.5 (range, 0–168). Four (20%) patients had0 CTC, eight (40%) patients had 1–4 CTC, and theother eight (40%) patients had ‡5 CTC. Of theprimary tumors, 16 (80%) were ER+ and 11 (55%)were PgR+, only three (15%) showed amplification ofHER-2⁄neu. Biopsies of metastatic lesions were avail-able in 15 of the 20 patients with recurrent disease(Table 1). Nine (60%) of the metastatic tumors wereER+, seven (47%) were PgR+, and three (20%)showed amplified HER-2⁄neu. Expression patterns ofER (kappa = 0.71; p = 0.004), PgR (kappa = 0.60;p = 0.02), and HER-2 (kappa = 0.61; p = 0.047) weresimilar in the primary and metastatic tumors.In addition to ER, PgR, and HER-2, we examinedthe expression of two other genes of interest, mamma-globin (MGB) and Notch-1 in CTC. MGB levels inprimary breast tumors inversely correlate with tumoraggressiveness and axillary node invasion (8). Notch-1can be aberrantly activated in human breast cancercells (9) and may regulate the self-renewal of tumori-genic stem cells. Therefore, Notch-1 may represent agenetic biomarker of a subset of breast cancer stem