Abstract
Background/Aim: Accurate preoperative localization is essential for successful breast-conserving surgery on non-palpable breast lesions. In recent years, Magseed has emerged as a non-wired localization approach with promising outcomes due to its flexibility and precision. This meta-analysis aimed to evaluate the efficacy and safety of Magseed localization on non-palpable breast lesions.
Materials and Methods: Ovid MEDLINE, CINAHL, Ovid EMBASE, and The Cochrane Library were searched from inception to February 2025, following PRISMA guidelines. The pooled mean and proportions were analyzed using a random-effects model. The review was registered with PROSPERO (CRD420250654940).
Results: From 958 studies screened, 16 studies involving 2,117 patients and 2,176 Magseeds were included. The overall rate of positive margins was 7.6% [95% confidence interval (CI)=0.04, 0.11, I2=88%], and the re-excision rate was 8.2% (95%CI=0.05, 0.12, I2=87%). The overall complication rate was 0.6% (95%CI=0.001, 0.011, I2=7%). The success rate for Magseed placement was 99.3% (95%CI=0.987, 0.998, I2=1%), the migration-related failure was 0.6% (95%CI=0.001, 0.011, I2=0%), and the retrieval success rate was 99.6% (95%CI=0.992, 0.999, I2=0%). The mean operative time was 61.4 min (95%CI=55.9, 66.9, I2=98%).
Conclusion: Magseed appears to be a safe and effective technique for the preoperative localization of selected non-palpable breast lesions. Nonetheless, further studies based on breast density, size, and depth are required to investigate the feasibility of preoperative Magseed localization for patients with non-palpable breast lesions.
Introduction
Breast cancer remains a major global health challenge, accounting for 12.5% of all newly diagnosed cancer cases worldwide in 2020, with over 2.3 million individuals affected annually (1). Advanced screening programs have enhanced early detection rates, resulting in approximately 30% of breast cancers being diagnosed as non-palpable lesions (2). For these lesions, breast-conserving surgery (BCS) has become the cornerstone of treatment, prioritizing oncological efficacy while preserving breast aesthetics and patient quality of life (3). However, the success of BCS hinges on accurate preoperative localization of lesions to ensure complete tumor excision with minimal resection of surrounding healthy tissues (3).
BCS with wire-guided localization (WGL) was first described in 1965 by Dodd et al., and it has been the standard surgical approach for non-palpable breast lesions (4). This technique consists of inserting a thin wire into breast tissue under imaging guidance to accurately localize non-palpable lesions prior to surgery (5). Nonetheless, WGL requires wire placement and excision of lesions on the same day, leading to scheduling challenges. Furthermore, Davey et al. reported that WGL could result in high wire migration rates, which contribute to suboptimal excision margins and reoperation rates of up to 18.2% (6).
To overcome the limitations of WGL, the Magseed localization technique was first introduced by Douk et al. in 2013 as a promising alternative to WGL (7). The Magseed marker is a 5-mm non-radioactive magnetic titanium seed, and it received approval from the U.S. Food and Drug Administration (FDA) in 2016 (8). The seeds could be implanted with the aid of ultrasonography or mammography weeks before BCS (9), offering great scheduling flexibility. This is particularly beneficial for patients undergoing neoadjuvant chemotherapy to downstage tumors preoperatively (10, 11). Additionally, D’Angelo et al. observed that lesions with Magseed localization could be resected with smaller incisions using real-time intraoperative guidance via a handheld magnetic probe, and it could improve cosmetic outcomes (12).
Although the Magseed localization technique has been FDA-approved, questions regarding the success rates and postoperative complication rates of BCS with Magseed for the management of non-palpable breast lesions remain unanswered. To our knowledge, no existing single-arm meta-analysis has evaluated the clinical outcomes of the preoperative localization with Magseed. Thus, this study aims to evaluate the safety and efficacy of BCS with preoperative Magseed localization for the management of non-palpable breast lesions.
Materials and Methods
Search strategy and data sources. This review adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (13). A comprehensive literature search was conducted across multiple electronic databases, including Ovid MEDLINE(R), CINAHL, Ovid EMBASE, and The Cochrane Library, from their respective inception dates to the 23rd of February 2025. The search strategy was designed and conducted by an experienced librarian with input from the study’s principal investigator. Controlled vocabulary, supplemented with relevant keywords, was used to identify studies examining the usage of Magseed in the localization of non-palpable breast lesions. The complete search strategy, detailing the specific terms used and their combination within each database, is available in Supplementary Item 1. This systematic review was prospectively registered with PROSPERO (ID: CRD420250654940).
Eligibility criteria and quality assessment. Studies eligible for inclusion were randomized controlled trials and cohort studies involving adult patients (≥18 years) with non-palpable breast lesions undergoing preoperative Magseed localization for BCS. To be included, studies had to report at least one of the following outcomes: placement accuracy, retrieval success, margin positivity, re-excision, or overall complications. Case reports, case series, abstracts, conference abstracts, and articles that were not reported in English were excluded from the study. Studies were included only if they assessed Magseed localization for BCS, with a clearly defined therapeutic intent, as per the American Society of Breast Surgeons definition (14). Studies were excluded if Magseed was used in anatomical sites other than the breast, applied for total mastectomy or diagnostic excisional biopsy, or if they included any subset of palpable breast lesions. When two studies overlapped, we retained the one with higher methodological quality and the more comprehensive dataset. Article screening and extraction were conducted twice by two of seven independent assessors The quality of each study was assessed independently by five authors (ARA, EJN, LMD, MYH, RMA) using the risk of bias in non-randomized studies - of interventions (ROBINS-I) (15). Any disagreements were adjudicated by HN and discussed with co-authors as necessary.
Extracted outcomes and definitions. Extracted baseline characteristics included the type of breast lesion, histopathology, dimensional parameters, localization method, and the number of seeds inserted. The following outcomes were extracted: rates of placement success, retrieval success, margin positivity, re-excision, operative time, overall complications, and the interval between seed placement and surgery. High-grade etiology was extracted for lesions when dual pathologies were reported. In this study, the number of seeds was defined as the total number of Magseeds placed within a specific anatomical region. A lesion was defined as an abnormal growth within the breast tissue. Placement success was defined as Magseed positioned within 10 mm of the lesion; any displacement greater than 10 mm attributable to seed migration was considered a migration-related failure. All failures, including those due to migration and technical access issues, were included in the overall failure analysis (16).
Statistical analysis. The pooled means and proportions of our data were analyzed using a random-effects, generic inverse variance method of DerSimonian and Laird, which assigns the weight of each study based on its variance (17). The heterogeneity of effect size estimates across the studies was quantified using the Q statistic and I2. A value of I2 of 0-25% indicates insignificant statistical heterogeneity, 26-50% low heterogeneity, and 51-100% high heterogeneity (18). Furthermore, a leave-one-out sensitivity analysis was conducted to assess each study’s influence on the pooled estimate by omitting one study at a time and recalculating the combined estimates for the remaining studies. Publication bias was assessed using a funnel plot (19). This study followed the intention-to-treat analysis. If mean and standard deviation (SD) were unavailable, the median was converted to mean using the formulas from the Cochrane Handbook for Systematic Reviews of Interventions (20). Data analysis was performed using Open Meta analyst software (CEBM, Brown University, Providence, RI, USA).
Results
Study selection and patient characteristics. The initial literature search of the electronic databases yielded 957 studies. After removing duplicates, the articles were screened for inclusion and exclusion criteria, and 119 studies were retained for full-text review. Sixteen unique studies involving 2117 patients were included in this meta-analysis. Of the sixteen studies, ten (3, 11, 12, 16, 21-26) were retrospective, and six (27-32) were prospective cohort studies. Fourteen of the included studies (3, 11, 12, 16, 21-27, 29, 30, 32) were single institutional studies, while two studies (28, 31) were performed in multi-center settings. The mean age of patients ranged from 51.9 to 64.0 years. Of the seven studies reporting on sex (11, 16, 21, 22, 24, 25, 28), there were 1,319 females (100%). A PRISMA flowchart of the study selection process is depicted in Supplementary Item 2. The baseline characteristics of the included studies are described in Table I.
Baseline characteristics of included studies.
Risk of bias and quality assessment. Utilizing ROBINS-I tool for quality assessment, ten retrospective cohort studies were found to be of serious risk of bias and quality (3, 11, 12, 16, 21-26). Serious risk concerns were noted in domains of bias due to confounding (3, 11, 12, 16, 21-26), and bias due to deviations from intended intervention (16). Of the prospective cohort studies judged, one was found to be of moderate risk (32). Five were found to be of serious risk (27-31). Serious risk concerns were noted in domains of bias due to confounding (27-31), and bias due to missing data (28). Nonetheless, all studies were deemed eligible for analysis as patients appeared to represent the whole experience of the investigator. The exposure and outcomes were adequately ascertained, and the lengths of follow-up were adequate for the purposes of this study. Results of the quality assessment of all included studies are shown in Supplementary Item 3.
Clinical characteristics. A total of 2117 patients underwent 2176 Magseed placements. The pooled mean time between Magseed placement and surgery was 4.6 days [95% confidence interval (CI)=0.5, 8.6, I2=94%, n=340] according to four studies (11, 24, 26, 27). The total number of Magseeds placed per breast and laterality was assessed in fifteen studies with a sample size of 2073 Magseed placements (3, 11, 12, 16, 21-31). One Magseed was placed per breast for 1970 (95.0%) Magseed placements (3, 11, 12, 16, 21-31), whereas the use of multiple Magseeds for bracketing several lesions in a breast was reported in 103 (5%) Magseed placements (11, 16, 23, 27, 29). Unilateral placement was reported in 2065 (99.6%) Magseed placed (3, 11, 12, 16, 21-31), whereas 8 (0.4%) were placed bilaterally (25).
Among twelve studies (3, 11, 12, 16, 23, 25-29, 31, 32), 1847 Magseed placements were localized either via ultrasound (n=1,182, 64%) or mammogram (n=665, 36%). Of the eleven studies (3, 12, 22-25, 27, 29-32) reporting lesion size, the pooled mean size was 11.7 mm (95%CI=10.0, 13.4, I2=91%) across 1077 lesions. Among thirteen studies (3, 11, 12, 21-27, 30-32), the respective histopathology was reported as invasive cancer in 717 (66%) lesions, 173 (16%) as ductal carcinoma in situ (DCIS), 70 (7%) as benign, 57 (5%) as unspecified malignancy, 33 (3%) as high-risk lesions, and 33 (3%) as unspecified. Lesions were categorized based on type across three studies (11, 26, 29). Lesions were classified as a mass for 171 (69%) lesions, 70 (28%) with calcification, three (1%) with architectural distortion, and six (2%) with other. Lastly, among ten studies (12, 22, 23, 25-29, 31, 32), neoadjuvant chemotherapy was given in 93 (6%) lesions. Clinical characteristics are summarized in Table II.
Clinical characteristics of lesions and corresponding Magseed placement.
Perioperative outcomes. The operative time was reported in five studies (23, 24, 26-28). The pooled mean operative time was 61.4 min (95%CI=55.9, 66.9, I2=98%). Fourteen studies (3, 11, 12, 16, 21-23, 25-27, 29-32) evaluated the successful retrieval rate of Magseed, and the pooled successful retrieval rate was 99.6% (95%CI=0.992, 0.999, I2=0%, n=1218) across 1226 Magseeds. Ten studies (3, 11, 12, 16, 23, 25, 26, 29, 30, 32) reported on the successful Magseed placement, and the pooled success rate was 99.3% (95%CI=0.987, 0.998, I2=1%, n=976) among 989 Magseeds. Failed placement rate was 0.7% (95%CI=0.002, 0.013, I2=1%, n=13), with a migration-related failure rate of 0.6% (95%CI= 0.001-0.011, I2=0%, n=9) involving 989 Magseeds across ten studies (3, 11, 12, 16, 23, 25, 26, 29, 30, 32). Three studies (11, 29, 32) reported 329 Magseed placed at a distance less than 5mm from the lesion itself (97%, 95%CI=0.95, 0.99, I2=0%, n=329) across 342 Magseeds. The perioperative outcomes of the included studies are depicted in Figure 1.
The pooled estimate of perioperative outcomes. A) Mean operative time (min); B) Successful placement; C) Failed placement; D) Migration-related failure; E) Successful retrieval; F) Magseed placed less than 5 mm from lesion.
Postoperative outcomes. Of twelve (3, 11, 12, 16, 21, 23-26, 29, 30, 32) included studies, the overall complication rate was 0.6% (95%CI=0.001, 0.011, I2=7%, n=14) involving 1190 Magseeds. Among overall complications, the most reported complications were ecchymoses (n=6), intraoperative seed dislodgment (n=5), hematomas (n=2), infection (n=1).
Thirteen studies (3, 11, 12, 21, 23-31) reported re-excisions with an overall re-excision rate of 8.2% (95%CI=0.05, 0.12, I2=86%, n=195) across 1789 lesions. Among fifteen studies (3, 11, 12, 21-32), the pooled estimate of a positive margin rate was 7.6% (95%CI=0.04, 0.10, I2=88%, n=203) involving 1889 lesions, and the close margin rate was 0.4% (95%CI=0.001, 0.007, I2=11%, n=15) across 1889 lesions. Three studies (11, 29, 31) reported on the histopathology of lesions with positive margins; all 27 lesions were either DCIS or invasive cancer. The post-operative outcomes are described in Figure 2.
The pooled estimate of postoperative outcomes. A) Positive lesion margins; B) Close lesion margins; C) Re-excision; D) Overall complications.
Discussion
Non-palpable lesions are detected more frequently with the rise in early breast cancer screening. Consequently, BCS with preoperative Magseed localization has gained interest since FDA approval for its precision and effectiveness. The primary aim of this meta-analysis was to investigate the safety and efficacy of BCS with preoperative Magseed localization for non-palpable breast lesions. Based on the sixteen included studies, preoperative localization with Magseed demonstrated a high placement success rate with a low migration-related failure rate. Additionally, BCS with Magseed localization resulted in low incidents of positive margins and re-excision rates, as well as overall complication rates.
This single-arm meta-analysis is the first to evaluate the success rate of the preoperative Magseed localization technique for the management of non-palpable breast lesions. Our study demonstrated that the Magseed localization technique could accurately locate non-palpable lesions with a high success rate. Association of Breast Surgery and The National Health Service (NHS) Breast Screening Programme guidelines have set a clinical standard for localization procedures, recommending that at least 95% of lesions should be localized within a 10 mm radius of the target to ensure accurate surgical excision (33). Although initially established for WGL, this threshold remains relevant as a clinical standard for assessing the efficacy of newer techniques (4, 34). Our analysis demonstrates that the success rate of Magseed exceeds this benchmark, achieving a placement accuracy of 99.3% within 10 mm, and notably, 96.6% within an even stricter 5 mm radius. Furthermore, Cartagena et al. argue that the enhancement in the localization accuracy could lead to lower positive margin rates and re-excision rates (35). Moreover, the high level of accuracy reduces the requirement for intraoperative margin shaves, thereby minimizing unnecessary resection of healthy breast tissue, a fundamental goal of BCS (35). This millimetric accuracy could set new clinical expectations for BCS for non-palpable breast lesions, and it could potentially influence future guidelines toward more stringent sub-5 mm localization standards. However, it’s important to note that while the marker’s position is precise, it does not delineate the tumor’s perimeter. To address this, Parisi et al. have proposed combining wireless localization with intraoperative ultrasound to better visualize the lesion’s margins and further improve the accuracy of the resection (36).
Our meta-analysis is the first to evaluate the migration rate of Magseed for non-palpable breast lesions, and our study demonstrated low migration rates among included studies. Hayes et al. observed that Magseed localization could be performed up to thirty days prior to surgery for low migration rates, and this technique is a noninferior alternative to traditional wire localization (37). In our study, the mean interval between placement and surgery was approximately five days, yet the migration rate was less than one percent. Cheang et al. argue that the limited possibility of displacement is the major advantage of the Magseed localization technique, allowing flexibility in operating room scheduling (38). Nonetheless, Singh et al. postulated that Magseed migration could occur after localization and is frequently associated with mammographically placed seeds (8). This phenomenon is known as the “accordion effect”, where compressed breast tissue recoils upon decompression (8). Consequently, this recoil can cause the seed to migrate along the needle tract in the opposite direction to the compression and away from the intended location. To mitigate the migration risk caused by the accordion effect, Kapoor et al. recommended applying minimal compression during seed placement and slow, incremental decompression (39). Adopting these strategies for Magseed localization can further reduce the already low migration risk and enhance its reliability for patients requiring extended implantation and operating room scheduling flexibility (10, 40). A streamlined clinical process is evident in the study by Tayeh et al., which highlighted both the scheduling flexibility afforded by low migration risk and a rapid deployment time averaging 3.4 min (41). Our meta-analysis observed a high intraoperative retrieval rate of Magseed even when bracketing is employed for extensive lesions. This finding was consistent with the systematic review by Gera et al. (42). In most cases, the average spacing between bracketed seeds adhered to the 2 cm guideline, which aims to enhance signal clarity and minimize interference (29). Compared to the wire-guided bracketing approach, Magseed bracketing is relatively easy and allows surgeons to localize each seed (34). Additionally, the Magseed technique minimizes the cumbersome and uncomfortable process of bracketing multiple tumor foci with wires. Depretto et al. conducted a subgroup analysis on the effect of breast densities, sizes, and depth on the retrieval rate, and they observed a similar success rate for the retrieval of Magseed (34). Conversely, Dehaene et al. reported difficulty in detecting the Magseed signal until incising and inserting a probe into the breast tissue on rare occasions (32). Another practical challenge is the need to use non-magnetic surgical instruments, as standard tools can cause undesirable signal interference (43). When intraoperative Magseed localization fails, alternative localization methods, such as WGL, are used since inserting a second Magseed risks signal interference (44). Thus, further studies stratifying breast lesions by density, size, and depth are warranted to more accurately determine the retrieval success rate and overall efficacy of the Magseed localization technique. Importantly, the ongoing MELODY study aims to evaluate various localization techniques including Magseed in over 7,000 patients undergoing BCS for invasive breast cancer or DCIS across 20 countries. By assessing outcomes such as margin status, re-excision rates, patient satisfaction, and economic impact, MELODY is expected to fill critical evidence gaps and guide future clinical decision-making in breast lesion localization (45).
Study limitations. First, the included studies were predominantly single-center and were observational in design, with no randomized controlled trials. This inherently introduces potential for selection and information bias. Second, many of the included studies were limited by small sample sizes and demonstrated a serious risk of bias, primarily due to inadequate control of confounding variables or missing data. Moreover, some outcomes reported as medians with interquartile ranges had to be converted to means and standard deviations, which may add imprecision. Furthermore, inconsistent definitions of positive margins, along with other methodological variations, complicated the direct comparison and synthesis of findings related to localization accuracy and clinical outcomes. Another limitation lies in the absence of standardized guidelines regarding acceptable performance thresholds for localization techniques. Future studies should adopt standardized benchmarks for defining migration distance and successful placement to facilitate meaningful pooled analyses. Additionally, there was considerable heterogeneity in the outcomes, such as positive margin and re-excision rates. Lastly, several studies failed to report key variables such as seed depth as well as breast density, size, and depth. This incomplete reporting underscores the need for future research to systematically document these parameters. Future investigations should explore how cancer subtype may affect outcomes in the context of Magseed localization to allow for a more precise assessment of its efficacy.
Conclusion
This meta-analysis demonstrated Magseed as a safe and effective preoperative localization technique for BCS in the management of selected non-palpable breast lesions. Notably, our study observed low rates of positive margins and re-excision rates, high placement accuracy, and a low incidence of complications. Importantly, while the data suggest promising surgical and clinical benefits, the absence of comparative trials in this study limits the comprehensive evaluation of Magseed localization technique in the scope of alternative preoperative localization techniques. Despite the promising results, further studies based on breast density, size, and depth are required to investigate the feasibility of preoperative Magseed localization for patients with non-palpable breast lesions to guide clinical decision-making and optimize patient care.
Acknowledgements
The Authors would like to thank Olivia Schaff from Manchester University Hospitals NHS Foundation Trust for the literature search.
Footnotes
Authors’ Contributions
CAT, HN, LMD, MA, SAA conceived and designed the study. ARA, AG, EJN, LMD, MA, MYH, RMA reviewed the literature, collected, analyzed and interpreted the data and drafted the manuscript. All Authors read and approved the final manuscript.
Supplementary Material
The supplementary material for this article is available at: https://github.com/Lameesmasoud/magseed-localization-supplementary/blob/main/Supplementary%20Item%20AR.docx
Conflicts of Interest
All Authors declare no conflicts of interest in relation to this study.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.
Artificial Intelligence (AI) Disclosure
No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.
- Received October 9, 2025.
- Revision received October 18, 2025.
- Accepted October 22, 2025.
- Copyright © 2026 The Author(s). Published by the International Institute of Anticancer Research.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.