Correlation Between Margin Status and Ipsilateral Breast Tumor Recurrence in Patients With Breast Cancer Undergoing Breast-Conserving Surgery With Whole-Breast Radiation Therapy.

INTRODUCTION
In patients with invasive breast cancer, breast-conserving surgery (BCS) followed by whole-breast radiation therapy has become the standard treatment to achieve complete tumor removal while minimizing local recurrence. The status of the surgical margins plays a crucial role in this context. Specifically, the concept of "no tumor on ink," meaning no cancer cells are present on the inked surface of the excised tissue, has been widely accepted as an essential criterion to reduce the risk of ipsilateral breast tumor recurrence (IBTR) [1]. Several studies have demonstrated that when surgical margins are free of tumor cells, the risk of IBTR is significantly reduced [23].
However, previous studies on IBTR did not consider the three-dimensional structure of the breast, which may have affected the accuracy of identifying and assessing IBTR [45]. While some studies suggest that superficial and deep resection margins (RMs) are insignificant, the actual significance of these margins remains uncertain, especially when the tumors are close to the chest wall or skin surface [6]. For instance, some studies have shown that deep margins may be associated with higher local recurrence rates, challenging the notion that these margins are universally insignificant [7].
Moreover, guidelines advocating the "no tumor on ink" margin were based on patient data collected before significant advancements in modern radiation therapy and systemic therapy [13]. Since then, advances in imaging techniques and radiation therapies have emerged, potentially altering the relevance of these historical guidelines [8]. In addition, there is evidence that modern systemic therapies, such as human epidermal growth factor receptor type 2 (HER2)-targeted therapy, significantly affect recurrence patterns, which were not accounted for in previous studies [910].
Therefore, we aimed to redefine IBTR by considering the three-dimensional structure of the breast, including the distance and location of recurrence. We also analyzed the superior and deep margins, which have not yet been clearly defined. Finally, we aimed to explore whether margin status continues to have clinical relevance in patients treated with contemporary adjuvant radiotherapy and systemic therapies.

METHODS
We included patients with breast cancer who underwent BCS between 2011 and 2015 at Samsung Medical Center (SMC) in South Korea. Patients who had metastases before surgery, received neoadjuvant therapy, did not undergo radiation therapy, or had synchronous and metachronous bilateral breast cancer were excluded. Additionally, patients with lobular carcinoma in situ, ductal carcinoma in situ (DCIS), phyllodes tumors, or other benign tumors were excluded, with a focus on patients with invasive breast cancer. In addition, patients with a follow-up period of less than two years were excluded, as the duration was insufficient to detect local recurrence, and including these cases could potentially lead to an underestimation of the IBTR rate. Patients with insufficient margin data and no residual tumors after excisional biopsy or vacuum-assisted biopsy (VAB) were also excluded.
A schematic diagram of patient selection process is shown in Figure 1. A total of 8,139 patients underwent breast surgery at SMC between 2011 and 2015. Of these, 5,306 patients underwent BCS. After applying the exclusion criteria, 3,712 patients were included in the study.
We defined IBTR in two ways. First, IBTR was defined as the occurrence of a new tumor within a 45-degree angle on either side of the original tumor or in the central portion. The central portion was defined as the area where the tumor center was within 3 cm of the nipple. This will be referred to as the “90-degree angle definition.” The 90-degree definition was initially derived from the conventional quadrant-based classification used in breast surgery. However, we also analyzed a 120-degree definition to account for clinically relevant recurrences that would otherwise have been excluded under a narrower angle. This exploratory classification aimed to better capture the anatomical relevance of the recurrence location and defined local recurrence as the occurrence of a new tumor within a 60-degree angle on either side of the original tumor or in the central portion. Figure 2 illustrates these definitions.
Angles were determined based on clockwise orientation described in breast magnetic resonance imaging (MRI) or breast ultrasonography (USG) reports. MRI was prioritized, and if the orientation was not clearly indicated on MRI, the direction provided by the USG was used as a reference. If neither imaging modality provided a precise clockwise orientation and only the quadrant direction was mentioned, the central direction of the quadrant was selected. For example, if only "left breast upper outer quadrant" was indicated, the direction was assigned to 2:30. Each hourly interval was set to 30°. Given the three-dimensional anatomy of the breast, when the primary tumor was located within 3 cm of the subareolar area, any recurrence occurring within a 3 cm radius, irrespective of its direction, was defined as IBTR based on our criteria.
Hormone receptor (HR) status, including estrogen receptor (ER) and/or progesterone receptor (PR) status, was determined by pathologists based on immunohistochemistry (IHC) results. Tumors were considered HR-positive if ≥ 1% of cells stained positively or if Allred scores were between 3 and 8 [11]. HER2 status was assessed using IHC and/or fluorescence in situ hybridization (FISH) or silver in situ hybridization (SISH). Tumors were classified as HER2-positive when the HER2/CEP17 ratio exceeded 2.0 in FISH or SISH testing, in accordance with the American Society of Clinical Oncology/College of American Pathologists guidelines [12].
In this study, superficial and deep margins refer to the anterior (skin-side) and posterior (chest wall-side) edges of the surgical specimen, respectively. These were analyzed separately from the radial margins (superior, inferior, medial, and lateral) in the assessment of margin status.
RM positivity was defined as the presence of ink on the radial margin of the final surgical specimen, regardless of the intraoperative frozen section results. Margin widths were reviewed based on data from histology reports. Margin status was assessed using two approaches. The superior, inferior, medial, and lateral margins were evaluated, excluding the superficial and deep margins. Second, an additional analysis was conducted in which superficial and deep margins were included along with other margins to assess their impact on recurrence. Margin status was categorized based on the distance: positive, close (less than 1 mm, excluding positive margins), 1 mm, 2 mm, and greater than 2 mm. The IBTR rate was analyzed for each margin status category. Patients with positive RMs who underwent re-excision were included in the analysis, with classification determined by the margin status of the re-excised specimen. There were no strict institutional guidelines for re-excision, and the decision was made at the operating surgeon’s discretion. In cases of invasive cancer involving the margin, re-excision was generally performed unless no residual tissue remained or if the patient declined further surgery. For DCIS, re-excision was sometimes omitted in cases that were considered focal without a clearly defined size threshold.
Finally, a two-proportion z-test was used to determine the margin length threshold that resulted in a statistically significant IBTR rate. We calculated the IBTR rate using both the “90-degree angle definition” and the “120-degree angle definition,” and referred to these as the IBTR90 and IBTR120 rate. Additionally, we computed the IBTR rates separately for cases in which superficial and deep margins were excluded and those in which they were included. All analyses were performed using SPSS (version 25.0; IBM Corp., Armonk, USA).
The study protocol was reviewed and approved by the SMC Institutional Review Board (IRB) (IRB number: 2024-03-071). The requirement for patient consent was waived due to the retrospective nature of the study and the use of anonymized data.

RESULTS

Table 1 presents a comparison of the clinicopathological characteristics of patients with and without IBTR. This analysis was conducted based on the overall incidence of IBTR without applying 90°or 120°angle-based definitions. Several factors showed statistically significant differences between the two groups. Patients in the IBTR group had a higher proportion of high-grade tumors (49.4% vs. 27.7%, p < 0.001). In terms of HR status, ER negative (42.7% vs. 20.9%, p < 0.001) and PR negative (55.1% vs. 28.0%, p < 0.001) tumors were significantly more frequent in the IBTR group. Similarly, HER2 positive disease was more prevalent in the IBTR group than in the non-IBTR group (36.0% vs. 17.0%, p < 0.001). Finally, the incidence of distant metastasis was significantly higher in patients with IBTR (11.2% vs. 4.5%, p = 0.007). Among the 3, 712 patients in the total cohort, the overall local recurrence rate was 2.4% (89/3,712), the regional recurrence rate was 1.2% (44/3,712), and the distant metastasis rate was 4.7% (174/3,712).

Table 2 shows IBTR90 and IBTR120 for each margin status. We analyzed the data separately for cases in which superficial and deep margins were either excluded or included. First, when excluding superficial and deep margins, 96 patients had positive RMs, with IBTR90 occurring in one patient, resulting in an IBTR90 rate of 1.0% (1/96). When the margin was less than 1 mm, the total number of patients was 219, with an IBTR90 rate of 3.2% (7/219), the highest observed rate. When superficial and deep margins were included, the positive margin group consisted of 156 patients, with an IBTR90 rate of 2.6% (4/156). For margins less than 1mm, there were 325 patients with an IBTR90 rate of 2.5% (8/325). A total of 44 patients with unspecified margin lengths were excluded. In IBTR120, when superficial and deep margins were excluded, the IBTR120 rate was 1.0% (1/96) for margin-positive cases and 4.6% (10/219) for close margins. Compared to IBTR90 (n = 7), the number increased to 10 under IBTR120, showing a notable difference in the close margin group. Next, when superficial and deep margins were included, the IBTR120 rate was 2.6% (4/156) for margin-positive cases and 3.1% (10/325) for close margins.

Table 3 compares IBTR rates using the 90-degree and 120-degree angle definitions across margin length categories with and without the inclusion of superficial and deep margins. Margin positivity was defined as A; < 1 mm, B; 1 mm, C; 2 mm, D; and > 2 mm, E. We compared groups A vs. B, C, D, E and A, B vs. C, D, and E. IBTR rates were calculated using raw counts from Table 3. For instance, under the IBTR90 definition, excluding superficial and deep margins, the recurrence rate for group A was 1 of 96 (1.0%), while the rates for groups B, C, D, and E were 65 of 3,616 (1.8%), as calculated by summing the respective counts across categories B, C, D, and E. Importantly, when superficial and deep margins were excluded, the IBTR120 rate in groups A and B group was 3.5% (11/315), compared to 1.8% (62/3,397) in the C, D, E group, showing a statistically significant difference (p = 0.042). However, this difference disappeared when superficial and deep margins were included in the margin status analysis, suggesting that these margins did not have a significant impact on the correlation between margin status and IBTR.
In addition, we identified 40 patients who had no residual tumors in the final surgical specimen after excisional biopsy or VAB. Among them, one patient (1/40, 2.5%) who had undergone VAB developed IBTR during follow-up. While this subgroup was not included in the margin-based analyses, the findings may raise concerns about the adequacy of initial biopsy procedures in cases of invasive carcinoma.

DISCUSSION
In this study, we redefined IBTR using anatomical criteria based on 90° and 120° angles, and analyzed its association with margin status. When superficial and deep margins were excluded, the recurrence rate under the IBTR90 definition was 3.2% (7/219) in the close margin group and 1.0% (1/96) in the positive margin group. Under the IBTR120 definition when superficial and deep margins were excluded, the recurrence rates in the close margin group were 4.6% (10/219) and 1.0% (1/96) in the positive margin group. Notably, when comparing the grouped categories (Table 3), the IBTR120 rate was significantly higher in groups A and B group (3.5%, 11/315) than in the C, D, E group (1.8%, 62/3,397; p = 0.042). These results suggest that close margins may be associated with a higher risk of recurrence within anatomically adjacent regions; however, unexpectedly, the recurrence rate in the positive margin group was not higher than that in the close margin group.
In contrast to previous studies on IBTR, our study introduces a new definition that considers both the direction and distance of recurrence. This approach considers the three-dimensional structure of the breast and its lobular architecture, which are critical for understanding tumor behavior and recurrence patterns. Traditional definitions often overlook these aspects, potentially missing finer details on how the breast anatomy influences recurrence. For instance, invasive lobular carcinoma, with its unique recurrence patterns, highlights the importance of considering the lobular structure of the breast in recurrence analyses [13]. Furthermore, the three-dimensional nature of the breast, as detailed in studies comparing different surgical approaches, suggests that a more nuanced understanding of tumor location and margins could improve recurrence prediction [1415]. The inclusion of breast anatomy, such as peritumoral vascular invasion and its prognostic implications, further supports the need for a comprehensive definition of IBTR that incorporates both anatomical and pathological considerations [16].
Many studies have examined the significance of superficial and deep margins in BCS and have often concluded that these margins do not have a substantial impact on recurrence rates. For example, Pleijhuis et al. [17] found that superficial and deep margins were not significantly associated with local recurrence in patients with early stage breast cancer undergoing BCS. Similarly, Yoon et al. [18] found that in most cases, the inclusion of superficial and deep margins did not correlate with improved outcomes, questioning the necessity of achieving wider margins in these areas. While these studies have demonstrated that superficial and deep margins do not significantly affect local recurrence rates, evidence suggests that close margins, including superficial margins, may influence outcomes under certain conditions. For example, a meta-analysis by the British Medical Journal found that margins less than 2 mm were associated with higher rates of local recurrence, underscoring the importance of adequate margin clearance in BCS regardless of the margin location [19]. Moreover, a study from British Journal of Surgery Open highlighted that, while complete excision with narrow margins may provide local control comparable to wider excision, the role of superficial and deep margins remains unclear. This ambiguity emphasizes the need for further research to definitively determine whether these specific margins truly contribute to local recurrence [20]. In our study, we included superficial and deep margins in the analysis to further explore their potential impact. However, our results did not show any significant association between margins and recurrence rates. This finding aligns with those of many previous studies, indicating that superficial and deep margins may not be as critical as previously thought. Further research is needed to determine the precise role of these margins in BCS and to establish guidelines for their consideration in clinical practice.
The current "no tumor on ink" guideline was developed based on studies conducted with patients treated before 2010, during which the IBTR rate was 5.3% [1]. However, our study focused on patients who underwent surgery between 2011 and 2015, a period that has seen significant advancements in both radiation therapy and systemic therapy compared to previous studies. At our institution, the overall IBTR rate was 2.4%, which was notably lower than those reported in previous studies. This decrease in IBTR rates may reflect improvements in modern radiation, hormone, and targeted therapies. Several studies have documented the impact of advancements in these therapies in reducing local recurrence rates. Offersen et al. [21] highlighted how modern radiation techniques, including hypofractionation and intensity-modulated radiation therapy, improve the precision and effectiveness of radiation, thereby reducing local recurrence. Curigliano et al. [10] discussed the role of updated hormone therapies, such as aromatase inhibitors and selective ER modulators, in decreasing the recurrence of HR-positive breast cancer. Similarly, the development of targeted therapies such as HER2-targeted agents has contributed to a decline in local recurrence rates by effectively managing residual disease and reducing tumor proliferation [910]. The better outcomes observed in our study may be due to these updated treatment strategies.
In our study, although no residual tumor was detected postoperatively following VAB or excisional biopsy, IBTR occurred at a rate of 2.5% (1/40). This IBTR rate was similar to the overall IBTR rate of 2.4% reported at our center. These findings suggest that performing VAB or excisional biopsy alone may not be sufficient in cases with potential for invasive cancer. Therefore, a more comprehensive surgical management should be considered to minimize the risk of recurrence. Similar studies have reported that even in cases where no residual tumor is found after biopsy, IBTR can still develop, further supporting the need for thorough surgical intervention and follow-up to prevent recurrence [2223]. Additional research supports the necessity of complete surgical excision following VAB in cases of invasive breast cancer. Park et al. [24] found that residual tumors were present in 81.6% of patients after VAB, underscoring the limitations of VAB as a definitive treatment method and highlighting the need for surgical excision to ensure adequate local control. Furthermore, excisional biopsy has limitations owing to its potential for incomplete tumor removal. A recent population-based study found that re-excision was required in 19% of patients after initial BCS, leading to an increased risk of recurrence and decreased survival rates compared with patients who achieved clear margins from the outset. This emphasizes the need for a comprehensive margin assessment and possible re-excision to minimize the recurrence risk and optimize outcomes [25].
This study had several limitations. First, tumor and breast sizes were not incorporated into the anatomical definition of IBTR. In cases where the tumor is large and the breast is small, the proposed angle-based definition may not have been applicable with precision. Second, the anatomical localization of recurrence relies on the clockwise orientation described in imaging reports, which can be inconsistent due to breast mobility depending on patient positioning. Third, patients in whom re-excision was performed for positive margins were excluded based on the surgeon’s clinical judgement. This subjectivity may have affected the analysis of margin status in relation to IBTR. Finally, given the retrospective and single-center nature of the study, potential selection bias and unmeasured confounders cannot be excluded. Despite these limitations, this study offers a novel exploratory approach for anatomical classification of IBTR in the modern treatment era.
In our study, when we excluded superficial and deep margins and applied the 120-degree angle definition to IBTR, grouping positive and close margins together and comparing them to other groups resulted in a statistically significant p-value of 0.042.
However, in all the other cases, no statistically significant differences were observed. This result differs from the current guidelines for invasive tumor RMs, which are based on the "no ink on tumor" principle. As previously mentioned, in cases of margin positivity, the surgeon's subjective judgement played a role, with re-excision being more commonly chosen when invasive cancer, rather than DCIS, was present at the margins. This may have influenced the incidence of IBTR. In addition, advances in modern radiation technology and systemic treatments may have contributed to these outcomes. In conclusion, when comparing IBTR rates based on margin length, re-excision should not be automatically chosen for a margin-positive state. Instead, the surgeon’s intraoperative assessment and judgement should be considered when deciding on re-excision.