Prognostic Significance and Predictive Value of Circulating Tumor Cells in Invasive Lobular Breast Carcinoma-An
Jose Luis Sandoval1,2,3, Nicolas Kiavue1, Lounes Djerroudi4
1Department of Medical Oncology, INSERM CIC 1428, Institut Curie, Paris, France.
Purpose:
In patients with metastatic breast cancer, ≥5 circulating tumor cells (CTC)/7.5 mL of blood is a validated adverse prognostic marker. The STIC CTC phase III trial (N = 778, hormone receptor-positive HER2- metastatic breast cancer) showed that ≥5 CTC/7.5 mL before treatment predicted greater benefit from first-line chemotherapy over endocrine therapy for progression-free (PFS) and overall survival (OS). This exploratory analysis examines these results by histologic subtype, comparing invasive lobular carcinoma (ILC) and carcinoma of no special type (NST).
Experimental Design:
Baseline CTC count (CellSearch) and its impact on PFS and OS were compared between ILC (N = 159) and NST (N = 571). The survival benefit of chemotherapy in patients with a high CTC count was then re-evaluated separately for each subtype.
Results:
Before treatment, ILC had significantly higher CTC counts than NST [median, 10 (Q1 = 2, Q3 = 40) vs. 1 (Q1 = 0, Q3 = 7)]. ≥5 CTC/7.5 mL were detected in 64% (95% confidence interval, 56-71) of patients with ILC and 31% (95% confidence interval, 28-35) of patients with NST, correlating with shorter PFS and OS in both. However, patients with ILC and ≥5 CTC/7.5 mL saw no significant benefit from CTC-informed chemotherapy [PFS: HR, 0.91 (0.55-1.52); OS: HR, 0.83 (0.46-1.52)]. In contrast, patients with NST and ≥5 CTC/7.5 mL had markedly improved outcomes with chemotherapy [PFS: HR, 0.54 (0.37-0.81); OS: HR, 0.37 (0.21-0.66)].
Conclusions:
ILC sheds more CTC than NST, likely owing to underlying biological differences. Although the ≥5 CTC/7.5-mL threshold remained a valid prognostic marker for both, it was predictive of chemotherapy benefit in NST but not in ILC. These findings highlight differences in biomarker utility between ILC and NST, affecting both threshold relevance and treatment response.
