Magnetic resonance imaging as a non-invasive breast cancer tool - the relevance of breast edema score: A retrospective study

Abstract

BACKGROUND

Breast magnetic resonance imaging (MRI) provides valuable information for tumor detection, as well as potential applications in molecular characterization and prognostication. A key feature detectable on MRI is the presence of breast edema, which has been associated with tumor aggressiveness and poorer clinical outcomes.

AIM

To determine the correlation between intramammary edema patterns as observed on breast MRI and the histopathological and molecular characteristics of the tumor.

METHODS

In this retrospective single-center study, 123 women with biopsy-proven breast cancer underwent preoperative breast MRI from June 2022 to June 2025. The classification of edema was determined on T2-weighted images, divided into four breast edema score (BES) categories: BES-1 (no edema), BES-2 (peritumoral edema), BES-3 (prepectoral edema), and BES-4 (subcutaneous edema). The MRI findings were correlated with histological type, molecular subtype, receptor status, Ki67 index, and lymph node involvement. Data analysis was conducted using IBM SPSS Statistics for Windows, version 28.

RESULTS

Edema was observed in 45.5% of patients. Statistically significant correlation was observed between BES and molecular subtypes (P < 0.001), hormone receptor status (P < 0.001), and human epidermal growth factor receptor 2 expression (P < 0.001). Higher BES categories (BES 2-4) exhibited a higher prevalence in human epidermal growth factor receptor 2-positive and triple-negative tumors, while the absence of edema (BES-1) demonstrated a predominance in hormone receptor-positive subtypes.

CONCLUSION

The presence and severity of MRI-based breast edema score have been found to correlate with aggressive molecular subtypes, underscoring the potential role of BES in prognostic stratification and guiding tailored treatment strategies.

Key Words: Breast magnetic resonance imaging; Breast cancer; Intramammary edema; Breast edema score; T2-weighted imaging; Molecular subtype; Hormone receptor status; Human epidermal growth factor receptor 2 expression; Ki-67 proliferation index; Prognostic imaging biomarker

Core Tip: Magnetic resonance imaging-based assessment of intramammary edema may serve as a practical imaging biomarker reflecting breast cancer biology. Using a simple four-level breast edema score on T2-weighted images, edema patterns can be linked to key histopathological and molecular features, including tumor subtype, receptor status, proliferation index, and nodal involvement. Higher breast edema score categories tend to accompany more aggressive molecular profiles, whereas the absence of edema is more frequently observed in hormone receptor-positive disease. Therefore, incorporating breast edema scoring into routine preoperative magnetic resonance imaging interpretation could support risk stratification and help guide individualized treatment planning in clinical practice.

INTRODUCTION

Breast cancer remains the most frequently diagnosed malignancy among women globally. In 2022, an estimated 2.3 million new cases were recorded worldwide, accounting for 23.8% of all cancer cases in women[1]. Incidence rates continue to rise, especially in industrialized countries. The disease poses a significant threat to both public health and socioeconomic systems. Early detection is crucial for improving survival and patient outcomes. Characterization of prognostic and predictive markers enables clinicians to create effective therapeutic strategies.

The most sensitive imaging modality for breast cancer detection is magnetic resonance imaging (MRI), outperforming mammography and ultrasound[2-4]. Although histological verification via biopsy remains indispensable, MRI contributes beneficial, additional information- enhanced visualization of tumor characteristics and identification of associated factors relevant to prediction and prognosis[2].

Breast edema, which is detected by MRI, is a clinically important radiological finding. Its presence and degree are often associated with more aggressive tumor subtypes. It is related to poorer prognosis and can indicate a higher risk profile[5-7]. Non-invasive evaluation of tumor biology and potential aggressiveness is enabled via such an advanced imaging modality. Verifying breast edema at an early stage may guide clinicians to opt for more intensive therapeutic interventions, which eventually may improve clinical outcomes.

This study aims to investigate the association between intramammary edema appearance on breast MRI and the histopathological and molecular profile of the tumor. The distribution of different breast edema score (BES) categories was specifically evaluated, and their associations with histological and molecular subtypes, receptor status, tumor grade, proliferative index, and lymph node involvement were analyzed.

MATERIALS AND METHODS

Study design

This is a retrospective, single-center, present study, which includes 123 women who were diagnosed with breast cancer through either an excisional or Tru-Cut biopsy from June 2022 to June 2025. All women underwent preoperative MRI at a minimum of 6 weeks following the biopsy procedure. In accordance with the current guidelines, all MRI examinations of premenopausal women were conducted within the timeframe spanning from the seventh to the fourteenth day of the menstrual cycle. The inclusion criteria comprised patients over the age of 18 years who had undergone a breast MRI and had a proven diagnosis of breast cancer, as confirmed by histology and immunohistochemistry. Exclusion criteria include cases within complete clinical, imaging, and pathological data, patients who also underwent neoadjuvant chemo therapy before the MRI, and examinations within adequate image quality. In addition, cases with clinical or imaging findings suggestive of alternative causes of edema, such as inflammatory or infectious processes or systemic conditions, were excluded. This approach was adopted to reduce potential confounding and to ensure that the analyzed edema reflected tumor-associated changes.

MRI acquisition protocol

The MRI examinations were conducted using a 1.5 T MRI scanner (Magnetom Amira, Siemens Healthcare, Erlangen, Germany), an 18-channel dedicated breast coil, and the women were positioned prone, with the head first. The examination encompasses the following sequences: A non-fat-suppressed T1-weighted transversal sequence; a fat-suppressed turbo inversion recovery magnitude (TIRM) transversal sequence, diffusion-weighted images with apparent diffusion coefficient maps, and axial dynamic fat-suppressed contrast-enhanced 3D T1 with subtraction images.

In order to evaluate the breast edema, the following sequences with their respective parameters were analyzed: Fat-suppressed TIRM transversal sequence - field of view 340/100 mm, time of repetition 3000 milliseconds, time of echo 55 milliseconds, matrix size 336/448, slice thickness 3 mm, and scanning time 156 seconds. Axial dynamic fat-suppressed contrast-enhanced 3D T1 - weighted fast low-angle l sequence-field of view 240/150 mm, time of repetition 4.35 milliseconds, time of echo 1.76 milliseconds, matrix size 384/256, slice thickness 1.1 mm, and scanning time 398 seconds.

Five contrast series were obtained - one pre-contrast and four post-contrasts. The contrast media utilized in this procedure were Gadovist 1 mmol/mL (a gadolinium-based contrast agent), with an administration rate of 3 mL/second, injected through a peripheral venous access, followed by 15 mL of saline.

MRI image analysis

MRI assessments were conducted by two radiologists. One radiologist had five years of experience in breast imaging, and the other had 15 years of experience. Both radiologists worked independently from one another and were unaware of the histopathological results. In cases of discrepant interpretations, a consensus review was performed to reach a final agreed score.

Regions with edema were assessed by comparison of TIRM and were compared with post-contrast T1-weighted images. On the TIRM sequence, edematous changes appeared as hyperintense areas, indicative of increased interstitial fluid. Those zones lacked enhancement after contrast administration.

Based on the existence and type of edema the patients were subdivided into four categories: (1) BES-1-without edema (Figure 1); and (2) BES-2 with peritumoral edema (Figure 2), BES-3 with prepectoral edema (Figure 3) and BES-4 with subcutaneous edema (Figure 4). If two types of edemas were found in one patient the one with the higher score was taken into account.

Figure 1 A 71-year-old woman with right breast invasive ductal carcinoma-no special type, luminal A breast cancer. A: Axial turbo inversion recovery magnitude image shows no edema; B: Axial subtraction images.

Figure 2 A 56-year-old woman with right breast invasive ductal carcinoma-no special type, triple-negative breast cancer. A: Axial turbo inversion recovery magnitude image shows peritumoral edema without contrast enhancement on; B: Axial subtraction images.

Figure 3 A 57-year-old woman with right breast invasive ductal carcinoma-no special type, triple-negative breast cancer. A: Axial turbo inversion recovery magnitude image shows prepectoral edema without contrast enhancement on; B: Axial dynamic fat-suppressed contrast-enhanced 3D T1-WI.

Figure 4 A 36-year-old woman with right breast invasive ductal carcinoma-no special type, luminal B, human epidermal growth factor receptor 2-positive breast cancer. A: Axial turbo inversion recovery magnitude image shows diffuse subcutaneous edema without contrast enhancement on; B: Axial subtraction images.

Furthermore, an evaluation of the lymph nodes was conducted, encompassing the assessment of their morphology, namely the presence of a round shape, the thickness of the cortical layer, and the absence of the conventional fat-equivalent center within the lymph node. The presence of any of these abnormalities was deemed indicative of metastatic disease.

Histopathological and molecular analysis

The following data were collected for the study: Histologic type, histologic grade, tumor grading, and lymph node status. Immunohistochemistry was also employed, including estrogen receptor, progesterone receptor, human epidermal growth factor receptor 2 (HER2), and Ki-67 index, which were used to classify tumors according to the 2013 St. Gallen International Expert Consensus Guidelines for Immunohistochemistry in breast cancer. The cut-off value of the Ki67 index was determined to be 20%. All histopathological and immunohistochemical analyses were performed as part of the routine clinical diagnosis in a certified pathology laboratory, following institutional protocols and internationally accepted guidelines.

Statistical analysis

Data analysis was conducted using IBM SPSS Statistics for Windows, version 28.0. Age distribution, being the only continuous variable, showed a lack of normality, and the central tendency was described with the median and interquartile range (interquartile range) and compared among BES lesion categories using the Kruskal-Wallis test. The remaining data consisted of categorical or ordinal variables, which were presented as n (%). Z-tests were employed to compare category proportions. To establish associations between variables, Fisher’s exact test was utilized for dichotomous data. For pairwise comparisons involving variables with multiple categories, the χ² test and Z-test with Bonferroni adjustments were employed. When expected counts in certain cells were less than five, the Monte Carlo method was applied with the χ² test. Statistical significance was accepted at P < 0.05. The authors have read the STROBE statement - checklist of items, and the manuscript was prepared and revised according to the STROBE statement- checklist of items.

RESULTS

Background information about the patients

The study included 123 female patients diagnosed with breast cancer, with a median age of 47 years (interquartile range = 20), an age range of 26-80 years, and the age group 41-60 constituted a significant majority (P < 0.001). Luminal B (HER2 negative) and luminal A were the two most prevalent molecular subtypes of breast cancer, whereas invasive ductal carcinoma was the dominant histological type (89.40%) against all other types (P < 0.001). According to tumor grading, grade II was significantly more prevalent than grade III and grade I (P < 0.001). The majority of the patients were estrogen (estrogen receptor) and progesterone (progesterone receptor) positive (P < 0.001 for both). HER2 receptors 1+ were detected in the majority of cases, followed by 0+ and 3+. Patients with a high proliferative index (≥ 20%) dominated significantly over those with a low proliferative index (< 20%), P < 0.001. Edema was observed in 45.50%, among which peritumoral edema (BES-2) was the most frequently present, vs prepectoral edema (BES-3) and subcutaneous edema (BES-4) (P < 0.001). Lymphadenopathy was detected in 56.30% of the cases (P = 0.030) (Table 1).

Table 1 Background information about the patients.

Variables	Statistics, n (%)	P value
Age	4
Median (interquartile range)	7 (20)
Minimum-maximum	26-80
Age categories
≥ 40 years	29 (23.60)a
41-60 years	61 (49.60)b	< 0.001
61-80 years	33 (26.80)a
Molecular subtype
HER2+	11 (8.90)a	Luminal B vs luminal A: P = 0.02; all others: P < 0.001
Luminal A	35 (28.50)b
Luminal B (HER2+)	52 (42.30)c	Luminal A vs HER2+, luminal B, and triple-negative breast cancer: P < 0.001
Luminal B (HER2-)	9 (7.30)a
Triple negative (triple-negative breast cancer)	16 (13.00)a
Histological type
Invasive ductal carcinoma	110 (89.40)a
Invasive lobular carcinoma	8 (6.50)b
Medullary carcinoma	2 (1.60)b
Invasive papillary carcinoma	1 (0.80)b
Invasive tabular carcinoma	1 (0.80)b
Mucinous carcinoma	1 (0.80)b	IDC vs each of the other types: P < 0.001
Tumor grading (G)
Grade I	5 (4.10)a	GI vs GII: P < 0.001
Grade II	84 (68.30)b	GI vs GIII: P < 0.001
Grade III	34 (27.60)c	GII vs GIII: P < 0.001
Estrogen status (estrogen receptor)
Positive (> 1)	96 (78.00)
Negative (0)	27 (22. 00)	< 0.001
Progesteron status (progesterone receptor)
Positive (> 1)	90 (73.20)
Negative (0)	33 (26.80)	< 0.001
Human epidermal growth factor receptor 2-receptors
0+	49 (39.80)a
1+	54 (43.90)a	< 0.001
3+	20 (16.30)b
Proliferative Ki67-index
< 20	35 (28.50)
≥ 20	88 (71.50)	< 0.001
Edema
Yes	56 (45.50)
No	67 (54.50)	0.160
BES
BES-1 (no edema)	67 (54.50)a
BES-2 (peritumoral edema)	37 (30.10)b
BES-3 (prepectoral edema)	7 (5.70)c
BES-4 (subcutaneous edema)	12 (9.80)c	< 0.001
Lymphadenopathy
Yes	70 (56.90)
No	53 (43.10)	0.030

^aP < 0.05.

^bP < 0.01.

^cP < 0.001.

HER2: Human epidermal growth factor receptor 2; BES: Breast edema score; IDC: Invasive ductal carcinoma.

Histopathological types vs the presence or absence of edema

There was no significant association between the presence or absence of edema and the histological types of breast cancer (P = 0.275). Invasive ductal carcinoma was the most prevalent histological type in both cases with edema (87.50%, n = 49) and those without edema (91.00%, n = 61). Invasive lobular carcinoma was identified in 8.90% (n = 5) of the cases with edema and in 4.50% (n = 3) of those without edema. Medullary carcinoma was found in 3.60% (n = 2) of the edema cases and was not observed in cases without edema. Invasive papillary carcinoma, invasive tubular carcinoma, and mucinous carcinoma were absent in edema cases and present in 1.5% (n = 1) of those without edema (Figure 5).

Figure 5 Histological types of breast cancer vs presence or absence of edema. BC: Breast cancer.

Comparison of clinical-pathological variables across BES categories

No significant age difference was found among the four BES categories (P = 0.612). Molecular subtype showed a significant association with BES categories (P < 0.001). Luminal A and luminal B (HER2-) were significantly more prevalent in lesions with BES-1. BES-2 was more frequent in luminal B (HER2-) and triple-negative breast cancer (TNBC). Lesions with BES-3 were mostly characterized by luminal B (HER2+) and TNBC, whereas lesions with BES-4 were mostly distinguished by the absence of luminal A and the presence of all other subtypes.

Tumor grading (G) did not show a significant association with BES categories (P = 0.326). Grade II was most commonly observed in all BES, followed by grade III and grade I in frequency. Estrogen receptor and progesterone receptor expressions were significantly associated with BES (P < 0.001 for both estrogen receptor and progesterone receptor). Positive estrogen receptor and progesterone receptor expressions were prevalent in lesions without edema, BES-1.

HER2 receptors were also significantly related to BES categories (P < 0.001). The main differences were observed in relation to HER2 1+ receptors, which were significantly more frequent in lesions with BES-1, as well as in relation to 3+ receptors, which were more often identified in BES-3 and BES-4.

Ki67 levels greater than 20% were common across all BES categories, while lymph node involvement was more frequently observed in lesions with BES-3 and BES-4. However, no significant associations were identified between BES categories and the proliferative Ki67 index (P = 0.184) or lymphadenopathy (P = 0.130) (Table 2).

Table 2 Comparison of breast edema score categories, n (%).

Variables	BES-1 (n = 67)	BES-2 (n = 37)	BES-3 (n = 7)	BES-4 (n = 12)	P value
Age, median (interquartile range)	47 (20)	50 (21)	44 (29)	54.50 (22)	0.6121
Molecular subtype
HER2+	2 (3.0)a	6 (16.2)a	1 (14.3)a	2 (16.7)a
Luminal A	30 (44.8)a	4 (10.8)b	1 (14.3)b	0 (0.0)b
Luminal B (HER2-)	33 (49.3)a	14 (37.8)a	1 (14.3)b	4 (33.3)a
Luminal B (HER2+)	1 (1.5)a	2 (5.4)a	3 (42.9)b	3 (25.0)b
Triple negative	1 (1.5)a	11 (29.7)b	3 (42.9)b	3 (25.0)b	< 0.0012
Tumor grading (G)
Grade I	2 (3.0)	2 (5.4)	1 (14.3)	0 (0.0)
Grade II	51 (76.1)	22 (59.5)	4 (57.1)	7 (58.3)
Grade III	14 (20.9)	13 (35.1)	2 (29.6)	5 (41.7)	0.3262
Estrogen status (estrogen receptor)
Positive (> 1%)	64 (95.5)a	20 (54.1)b	5 (71.4)a,b	7 (58.3)b
Negative (0%)	3 (4.5)	17 (45.9)	2 (28.6)	5 (41.7)	< 0.0013
Progesteron status (progesterone receptor)
Positive (> 1%)	61 (91.0)a	18 (48.6)b	4 (57.1)b	7 (58.3)b
Negative (0%)	6 (9.0)	19 (51.4)	3 (42.9)	5 (41.7)	< 0.0013
Human epidermal growth factor receptor 2-receptors
0+	26 (38.8)a	18 (48.6)a	2 (28.6)a	3 (25.0)a
1+	38 (56.7)a	11 (29.7)b	1 (14.3)b	3 (33.3)b
3+	3 (4.5)a	8 (21.6)b	4 (57.1)b	5 (41.7)b	< 0.0012
Proliferative Ki67-index
< 20%	24 (35.8)	6 (16.2)	2 (28.6)	3 (25.0)
≥ 20%	43 (64.2)	31 (83.8)	5 (71.4)	9 (75.0)	0.1843
Lymphadenopathy
Yes	33 (49.3)	22 (59.5)	5 (71.4)	10 (83.3)
No	34 (50.7)	15 (40.5)	2 (28.6)	2 (16.7)	0.1303

^aP < 0.05.

^bP < 0.01.

^cP < 0.001.

¹Kruskal-Wallis test.

²χ² test.

³Fisher’s exact test.

Proportions differ significantly between categories of breast edema score at the 005 level if marked with different subscript letters. HER2: Human epidermal growth factor receptor 2; BES: Breast edema score.

DISCUSSION

In the context of modern medicine, diagnostic imaging has found its valuable place as a source of additional essential information. In the management of breast diseases, while the pathological result remains paramount for patient treatment, advancements in imaging technology, such as MRI, have enabled the extraction of data regarding tumors, their surroundings, and the entire breast, all in vivo. At present, several predictive and prognostic biomarkers from the breast MRI related to duration of response, progression-free survival, and overall survival are documented in the extant literature. These factors include the presence of hypervascularization, feeding vessels, hyperintensity on T2, washout, central necrosis, rim enhancement, and peritumoral edema[8,9]. What’s more, the presence of peritumoral edema is reported as an independent prognostic factor for breast cancer[5,7,10].

In breast cancer, breast edema is the buildup of interstitial fluid caused by disruption of the balance between vascular permeability, lymphatic drainage, and tissue compliance due to malignant infiltration, tumor-associated angiogenesis, or treatment-induced changes in the breast’s vasculature and lymphatic architecture and function[11-13]. The different patterns of edema define different types of diagnostic and prognostic importance.

Peritumoral edema is a result of two main mechanisms: Increased vascular permeability and early lymphatic obstruction. Factors like vascular endothelial growth factor activate tumor angiogenesis, which causes the production of immature, leaky vessels that allow protein-rich plasma to leak into the stroma[5,8]. Meanwhile, by invading lymphovascular channels, the malignant cells impair drainage, which leads to the accumulation of localized interstitial fluid[5,14,15]. Trabecular edema is represented by thickening of the fibrous septa in the breast parenchyma. Drainage along these connective tissue pathways is compromised by the tumor infiltration or the compression of intramammary lymphatics. Diffuse parenchymal involvement is frequently associated with this type of edema, which reflects impaired regional lymphatic flow[14,16].

When superficial dermal lymphatics are blocked, usually because of tumor emboli or widespread inflammatory changes, subcutaneous edema occurs. It can also show up in advanced locally invasive disease, but it is typical of inflammatory breast cancer. The accumulation of protein-rich interstitial fluid in the skin and subcutaneous fat is caused by impaired lymphatic outflow. This results in the classic clinical sign - peau d’orange[15,17]. Skin thickening and an increase in signal in the superficial fat layer are imaging signs of subcutaneous edema[12,16]. Its presence indicates widespread lymphovascular spread and is strongly linked to a poor prognosis[5,17].

Prepectoral edema happens when enlarged metastatic nodes hinder fluid return along prepectoral pathways, or when malignant obstruction spreads into lymphatics draining the axillary region or chest wall. When fluid builds up in the tissues close to the pectoral muscles, it can be seen on MRI as a high signal intensity[12]. For staging and treatment planning, this finding has important ramifications as it implies advanced locoregional disease, involvement of the chest wall, or nodal metastasis[15,18].

According to the findings from the present study, there is no statistically significant difference in the presence of edema among various histological subtypes. This suggests that the development of breast edema is more closely related to tumor biology andmicro environment changes rather than to histological type alone. Similar to our results, other studies have also reported lower rates of peritumoral edema in invasive lobular carcinoma, presumably due to its distinctive growth pattern- although this association has not been conclusively demonstrated[5,6,19-21].

When comparing molecular subtypes, we found that the vast majority of tumors without associated edema belonged to luminal subtypes, comprising 44.8% luminal A and 49.3% luminal B HER2-negative carcinomas. Furthermore, a statistically significant association was identified between the presence of estrogen and progesterone receptor expression and the absence of edema. This finding suggests that tumors lacking edema on MRI are more likely to correspond to luminal, hormone receptor-positive subtypes, which are generally linked to more favorable prognostic profiles. Similarly, many other authors have reported a significant correlation between breast edema and non-luminal tumors[6,19,22]. A possible explanation is the typically slower growth kinetics and lower angiogenic potential of luminal tumors, which may lead to less impairment in lymphatic drainage and lower vascular permeability, thus reducing interstitial fluid accumulation.

Our study demonstrated that nearly all HER2- positives cases exhibited edema, further supporting the hypothesis that HER2-driven angiogenesis contributes not only to tumor growth but also to stromal changes that promote fluid accumulation visible on T2-weighted images. The receptor tyrosine kinase HER2 is part of the family of human epidermal growth factor receptors. Overexpression of HER2 has been shown to upregulate vascular endothelial growth factor, thereby facilitating neovascularization and increasing vascular permeability, processes that lead to increased extracellular fluid accumulation[6,11,22,23].

An association between breast edema and TNBC is also demonstrated. TNBC, characterized by the absence of estrogen receptor, progesterone receptor, and HER2 receptor expression, is well known for its aggressive clinical course and poor prognosis[24]. In our cohort, nearly all tumors of this molecular subtype presented with edema on MRI. A potential explanation may be the high Ki67 expression, which reflects increased proliferative activity, rapid tumor growth, and increased angiogenesis. Several authors, including Abdelbary et al[6], Uematsu et al[15], Panzironi et al[21], Huang et al[22], and Baltzer et al[25] have further reported a correlation between tumor size and the presence of edema. Although no statistically significant difference was identified between Ki67 expression and BES in our study, lesions with Ki67 values above 20% tended to present more frequently with edema. This trend might still be biologically relevant, as a high proliferative index reflects more aggressive tumor behavior, increased metabolic demand, and angiogenic signaling, all of which may promote vascular leakage and lymphatic drainage. The absence of statistical significance in our cohort is likely related to the limited number of cases within some BES subgroups, particularly in the higher edema categories, rather than the absence of a true biological association. In addition, the presence of lymphovascular invasion may contribute to this phenomenon by obstructing vessels and lymphatic drainage, increasing vascular permeability, and thus fluid extravasation[5,8,26,27]. Wei et al[28] identified peritumoral edema in 67.6% TNBC cases in comparison to none of the cases with fibroadenomas. Abdelbary et al[6] further demonstrated that all TNBC cases presented with peritumoral edema with a significantly higher prevalence of moderate (prepectoral) and severe (subcutaneous) edema. Similarly, Chen et al[12] reported an association between higher grades of edema and more biologically aggressive tumors.

The existing literature provides heterogeneous and conflicting evidence regarding the relationship between breast edema and lymph node involvement. While several studies have demonstrated a statistically significant association between edema and higher nodal stage[6,7,12,20], others have failed to confirm such a correlation[5,19]. In this study, lymphadenopathy was observed to be more prevalent in higher BES grades (3 and 4), although this did not reach statistical significance. This pattern may reflect progressive lymphatic obstruction and locoregional tumor spread, which are likely mechanisms behind the development of prepectoral and subcutaneous edema.

Overall, these findings suggest that breast edema seen on MRI reflects underlying tumor biology rather than being a nonspecific imaging feature. The observed associations with molecular subtype and hormone receptor status indicate that BES, as a structured method for grading the extent and distribution of edema, may serve as an indirect imaging biomarker of tumor aggressiveness and tumor-related microenvironmental alterations.

This study has several limitations. First, it is a single-center retrospective analysis, which may limit the generalizability of the findings. Second, the number of patients in certain subgroups was relatively small, potentially reducing the statistical power for subgroup comparisons. These limitations highlight the need for larger, prospective, multicenter studies to validate the observed associations.

CONCLUSION

MRI can be utilized to noninvasively assess the tumor microenvironment, thereby providing invaluable additional information. Routine evaluation and reporting of breast edema, including its BES, can be applied as an adjunct imaging marker to support preoperative risk stratification, to suggest the presence of more aggressive tumor biology, and to improve prediction of axillary nodal involvement and unfavorable molecular sub types. Thus, BES can complement established MRI descriptors and clinic pathological factors, potentially aiding treatment planning and supporting tailored therapeutic strategies.