Tumor Microenvironment of Metastasis Score Independently Predicts Distant Metastasis in Estrogen Receptor–Positive/HER2-Negative Breast Cancer
Tumor microenvironment of metastasis (TMEM) consists of direct contact between a macrophage, an endothelial cell, and a tumor cell. In a case-control study reported in the Journal of the National Cancer Institute, Rohan et al found that TMEM score was an independent predictor of distant metastasis in patients with estrogen receptor–positive/HER2-negative breast cancer.
Study Details
The case-control study was nested within a cohort of 3,760 patients with invasive ductal breast carcinoma diagnosed between 1980 and 2000 and followed through 2010. Case patients were those who developed distant metastasis. Control subjects were matched for age at and calendar year of primary diagnosis.
TMEM was assessed by triple immunostain, and IHC4 was assessed by standard methods. A TMEM consisted of each occurrence of a structure featuring direct contact between an invasive pan-Mena–expressing carcinoma cell, an endothelial cell, and a perivascular macrophage with no stroma between the tumor cell and the perivascular macrophage; the TMEM score was the total number of TMEMs observed in 10 high-power fields.
Of the 573 breast cancer tissue blocks examined, 481 (84%), representing 259 case-control pairs, were usable and were included in the study. On univariate analysis, tumor size, tumor grade, lymphovascular invasion, number of positive lymph nodes, Ki67, and hormone therapy were associated with increased risk of distant metastasis and estrogen receptor–positive or progesterone receptor–positive patients and those who had received chemotherapy had reduced risk. On multivariate analysis including these factors and TMEM score, only tumor size, nodal status, tumor grade, and treatment with hormone therapy were associated with risk of metastasis at P < .10, with these variables being retained in subsequent multivariate models.
Predictive Value
The median TMEM score in the case patients was higher than in control subjects (19 vs 14). In the total study population, there was a nonsignificant increase in risk of distant metastasis with increasing TMEM score after adjustment for clinical variables.
In multivariate analysis in patients with estrogen receptor–positive/HER2-negative disease, TMEM score was an independent predictor of distant metastasis (odds ratio [OR] = 2.70, P = .004 for trend, for highest vs lowest tertile), with an OR of 1.16 (95% confidence interval [CI] = 1.03–1.30) per each 10-unit increase. IHC4 score had a borderline positive association (OR = 1.06 per 10 unit increase, 95% CI = 1.00–1.13).
The significant association of TMEM score with risk persisted after adjustment for IHC4 score and clinical risk factors (OR = 2.67, 95% CI = 1.36–5.26, for highest vs lowest tertile). The area under the receiver operating characteristic curve for TMEM after adjustment for clinical variables in this population was 0.78. A TMEM composite score (defined as a linear equation of all variables in the multivariate model with the estimated coefficients from logistic regression as their coefficients) identified low-, intermediate-, and high-risk groups with absolute risks of distant metastasis of 5.9%, 14.1%, and 30.3%.
IHC4 score was not independently associated with risk of distant metastases in the total population and neither TMEM score nor IHC4 was significantly associated with risk in triple-negative disease or HER2-positive disease.
The investigators concluded, “TMEM score predicted risk of distant metastasis in [estrogen receptor–positive/HER2-negative] breast cancer independently of IHC4 score and classical clinicopathologic features.”
Thomas E. Rohan, PhD, of Albert Einstein College of Medicine, is the corresponding author for the Journal of the National Cancer Institute article.
The study was supported by Metastat, Inc.
The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.
