Contralateral parenchymal enhancement on dynamic contrast-enhanced MRI reproduces as a biomarker of survival in ER-positive/HER2-negative breast cancer patients

Breast cancer is a heterogeneous disease with inter- and intra-tumour heterogeneity. Different molecular subtypes have been defined that are associated with disease-free and overall survival (OS). Clinically, immunohistochemical surrogates are used to guide treatment including oestrogen receptor (ER)-positive/ human epidermal growth factor receptor 2 (HER2)-negative, HER2-positive and triple-negative [1‐4]. More recently, molecular assays have yielded additional tools to predict therapy outcome, helping to optimise strategies for patient-tailored treatment [5‐8]. Nonetheless, using current routine predictive markers in clinical practice, treatment outcome still varies within specific subtypes. Hence, an ongoing need for accurate risk stratification exists, where risk markers are correlated with outcome after breast cancer therapy.

Current routine tests mainly focus on the tumour; meanwhile, the breast parenchyma is relatively underexplored. Genomic changes in the parenchyma may generate cell transformations leading to malignancy [9]. Hence, the parenchyma may be important for breast cancer risk prediction, but also for treatment response and assessment of outcome [10‐17]. The parenchyma can be assessed using imaging and its relationship with patient outcome can be evaluated. Assuming symmetry between both breasts, one might hypothesise that the parenchyma in the disease-free breast (i.e. contralateral to the known cancer) is comparable to that in the diseased breast before tumourigenesis. Consequently, analysis of the parenchyma in the disease-free breast before treatment might gain insight into the role of the breast’s healthy parenchyma on patient outcome.

Earlier studies in patients with ER-positive/HER2-negative breast cancer showed that more pronounced contralateral parenchymal enhancement (CPE) on magnetic resonance imaging (MRI) prior to treatment was associated with improved patient survival, independent of other routine predictive markers [18, 19]. This effect was even more evident in patients receiving endocrine treatment. For proper assessment of the potential of CPE in patient stratification, this must be confirmed in an independent set of patients. Thus, the purpose of this study was to assess whether CPE, as an independent biomarker for patient survival, reproduces in an independent patient cohort from a different cancer institution.

This biomarker-assessment study shows that pre-treatment contralateral parenchymal enhancement (CPE) on MRI reproduces as a biomarker of long-term patient survival. In both the previous biomarker-discovery and the current biomarker-assessment study, patients with ER-positive/HER2-negative breast cancer with high CPE in the disease-free contralateral breast had significantly better invasive disease-free survival (IDFS) and overall survival (OS) than those with low CPE after adjustment for clinicopathological parameters and systemic therapy. In addition, CPE again showed complementary ability to improve the risk stratification of routine prognostic models. This study was performed in an independent population from a different cancer centre.

In the previous biomarker-discovery study, association between CPE and patient outcome was most evident in the group of patients receiving endocrine therapy. In the current biomarker-assessment study, most patients received endocrine therapy. Hence, comparison between the groups receiving endocrine therapy is most informative. Patients who received endocrine therapy with high CPE had a fourfold to fivefold better IDFS and OS than patients with low CPE after adjustment for clinicopathological characteristics in both studies. Note that this is regardless of the lower age at diagnosis and higher tumour grade in this biomarker-assessment study. Furthermore, a different MRI vendor was used (General Electric instead of Siemens) and differences in MRI acquisition existed; the current study used fat-suppression instead of no fat-suppression and anisotropic voxels instead of isotropic voxels. Therefore, a major strength of this study is that CPE reproduces across protocol (fat-suppression vs no fat-suppression, differences in flip angle, repetition time and voxel dimensions) and vendor (General Electric vs Siemens). More recently, CPE was found to improve on the risk stratification by clinicopathological and molecular-assay based models [19]. In the current study, we confirm the potential of CPE to complement the well-established NPI and PREDICT. Confirming the complementary value of CPE to molecular assays is subject of feature research.

The biological reason why pre-treatment CPE is associated with long-term survival is not yet fully understood. CPE is a measure of the perfusion of the healthy unaffected parenchyma. We tested previously raised hypotheses concerning hormone sensitivity in this paper: we previously hypothesised that tumours in patients with high CPE have a higher hormone sensitivity and, thus, may be more receptive to endocrine therapy [18]. However, we did not find any significant correlation between ER and PR-percentage of staining at pathology and CPE, just a non-significant trend with the ER-percentage. CPE did not show a correlation with the menstrual cycle, and CPE was not significantly higher in premenopausal women than in perimenopausal and postmenopausal women. Since CPE was not significantly associated with these measures of hormone activity, they do not explain the full extent of the correlation between CPE and survival. Although further investigation into the correlation between CPE and the menopausal status may be of interest for future research, in the current study we indirectly corrected for menopausal status in our survival analysis by adjusting for age.

Other hypotheses that are subject of future research to investigate why CPE is associated with long-term survival include facilitation of drug-transport and immune response. More pronounced CPE indicates a higher parenchymal perfusion. Hence, the ability of Tamoxifen to reach loco-regional targets might be increased. A way to test this is to compare the associations of CPE with survival in patients treated with Tamoxifen and in patients treated with aromatase inhibitors, since these two anti-hormonal therapies target different mechanisms [31, 32]. High immune response is associated with superior ability of the immune system to fight residual disease [33]. Such increased immune response may be correlated with high CPE (blood perfusion in the normal parenchyma). A potential candidate to measure immune response is the cytolytic activity signature, which is associated with counter-regulatory activities limiting immune response and an improved prognosis [33]. Testing these two hypotheses is subject of future studies.

Research on the enhancement of the parenchyma includes background parenchymal enhancement (BPE), the signal enhancement ratio (SER) and texture analysis. BPE relates to the parenchymal volume and intensity enhancing on MRI and is scored in four incremental categories [34]. Increased BPE has been associated with a higher risk of breast cancer development [11, 14, 35]. SER is the quantitative signal ratio between early and late subtraction scans [36]. Increased SER has been associated with longer disease-free survival in patients after one cycle of neoadjuvant chemotherapy [12]. Although our study is not aimed at response monitoring to neoadjuvant chemotherapy, it also shows a survival benefit in patients with higher perfusion of the parenchyma. It is likely that the underlying mechanism for this observation is, however, different. Investigating associations of CPE with patient survival in a neoadjuvant setting is of interest for future research. While Hattangadi et al. [12] focused on tumour-induced changes to the parenchyma directly around the index tumour, CPE focuses on the properties of the healthy parenchyma prior to tumourigenesis. Texture analysis of parenchymal tissue has been associated with signalling pathways and patient survival [37, 38].

Our study has some limitations. Firstly, we only included patients with invasive ductal carcinoma. Therefore, we cannot generalise to invasive lobular carcinoma or other invasive types. However, the biomarker-discovery study showed no difference between CPE and survival with respect to these histological tumour types [18]. Secondly, patients with biopsy-proven benign findings in the contralateral breast were excluded. Further algorithm development is necessary to ensure that these MRI findings are not in the segmentation. Thirdly, even though this study confirms CPE as a biomarker for survival, CPE values between institutions were systematically shifted. This is likely because of differences in vendor hardware and imaging protocols, and additional precautions are required when CPE is to be trichotomised in data pooled across institutions. Fourthly, we did not always perform the MRI in the recommended menstrual window. Postponing the MRI to correct for effects of the menstrual cycle would lead to undesired delay of surgery. This limitation was, however, present in the patients alive at follow up, patients with a recurrence, and patients who had died. Therefore, we consider this potential limitation unlikely to bias the results. Lastly, these were two retrospective studies. To fully elucidate the potential of the CPE biomarker, validation in a multi-cohort prospective study is desired.

To conclude, in this retrospective cohort from an independent cancer institution, contralateral parenchymal enhancement on pre-treatment dynamic contrast-enhanced MRI reproduces as an independent biomarker of survival in patients with ER-positive/HER2-negative breast cancer. These findings are a promising next step towards a practical, widely accessible and inexpensive test for risk stratification of ER-positive/HER2-negative breast cancer that can have a practice-changing impact.