Magnetic seed versus skin tattoo localization of non-palpable breast lesions: a single institution cohort study

Objective The objective of this retrospective study was to investigate the accuracy and feasibility of magnetic seed compared to skin tattoo in preoperative localization of impalpable breast lesions in terms of accuracy of placement, re-excision and positive margins rates, and breast/surgical specimen volume ratio. Methods We retrospectively analyzed 77 patients who underwent breast conservative surgery in our center from November 2020 to November 2021, with previous localization with skin tattoo or magnetic seed. Results Thirty-seven magnetic seeds were placed in 36 patients (48.6%) and 40 skin tattoos were performed in the remaining cases (51.4%). The seeds were placed correctly at the two-view mammogram acquired after the insertion in 97.6% (36/37) of cases. With both methods, 100% of the index lesions were completely removed and found in the surgical specimen. The reported re-excision rate was 0% for both groups. A significant difference was observed in the volume of breast parenchyma removed between the two groups, inferior in the seed group (p = 0.046), especially in case of voluminous breasts (p = 0.003) and small lesions (dimension < 8 mm, p = 0.019). Conclusions Magnetic seed is a non-radioactive localization technique, feasible to place, recommended in case of non-palpable breast lesions, saving the breast parenchyma removed compared with skin tattoo, without reducing the accuracy. Clinical relevance statement Our findings contribute to the current evidence on preoperative localization techniques for non-palpable breast lesions, highlighting the efficacy of magnetic seed localization for deep and small lesions. Key Points • Magnetic seed is a non-radioactive technique for the preoperative localization of non-palpable breast lesions studied in comparison with skin tattoo. • Magnetic seed is feasible to place in terms of post-placement migration and distance from the target lesion. • Magnetic seed is recommended in case of non-palpable breast lesions, saving the breast parenchyma removed without reducing the accuracy. Supplementary information The online version contains supplementary material available at 10.1007/s00330-023-10008-4.


Introduction
The constant improvement of imaging techniques and the introduction of mammography screening has led to an increased number of diagnosis of non-palpable breast cancers. Currently, about 30% of all breast cancers are not palpable at the time of diagnosis [1]. Non-palpable lesions cause an increased risk of positive tumor margins or excision of healthy parenchyma, resulting in high risk of local recurrence, poor cosmetic outcome and patient dissatisfaction [2]. Adequate preoperative tumor localization is mandatory to perform an appropriate conservative surgery, avoiding loco-regional recurrence and poor aesthetic outcomes. Nowadays, there are different techniques of localization: wire localization, carbon marking, radio-guided occult localization (ROLL), radioactive seed localization (RSL), non-radioactive radar localization, magnetic seed localization, intraoperative ultrasound and preoperative skin tattoo. The choice of which one to use depends on surgeon's and radiologist's experience, skills and available technologies. The most used method is the wire-guided localization (WGL) [3][4], however, this technique has several disadvantages [5,6,7].
In our Center, preoperative localization with skin tattoo is the preferred method by surgeon and radiologist for the excision of non-palpable breast tumors. In the last year, the magnetic seed localization (Magseed®, Endomagnetics, Cambridge, U.K.) has been introduced in our center. In 2016 the Food and Drug Administration (FDA) approved the rst magnetic seed marker with the SentiMag® localization probe [8][9] and has gained considerable interest. The magnetic seed is a 5x1mm stainless steel paramagnetic seed delivered in a sterile 18-G introducer needle [9]. The seed has no intrinsic magnetic activity. When inserted under ultrasound or stereotaxic guidance for the localization of non-palpable breast lesions, the seed is detected with a magnetic detection probe (Sentimag®) during surgery.
The aim of this retrospective single-center study was to evaluate the clinical safety and utility of the magnetic seed location system compared to skin tattoo in localization of impalpable breast lesions in terms of accuracy of initial placement, re-excision and positive margin rates and breast/surgical specimen volume ratio.

Patients
In this retrospective study we included 77 consecutive patients programmed to have breast conserving surgery with non-palpable breast lesions detected on clinical and/or screening mammography in the period between November 2019 and November 2020. In all cases, the preoperative histological diagnosis indicated breast cancers or B3 lesions [10].
Inclusion criteria were: age 18 or older, non-palpable unifocal lesion, pre-operative localization carried out with skin tattoo or magnetic seed localization, and surgery performed in our Institute. Exclusion criteria were: palpable lesion, patients underwent neoadjuvant chemotherapy, pre-operative localization conducted with different techniques from skin tattoo or magnetic seed and surgery performed in another Institute.
Each case has been subjected to multidisciplinary discussion with radiologists, breast pathologists and breast surgeons. During the discussion, imaging has been reviewed to de ne the best localization and treatment procedure. Data were retrospectively collected on patient demographics, cancer characteristics, and surgical details.
The volume of the mammary gland was calculated considering it as a cone ( 1 3 ϒ 2 h) from the mediolateral oblique view of staging mammogram [11]. The volume of the excised breast specimen was calculated as a geoid [12].
This study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Institutional Review Board of Fondazione Policlinico Universitario Agostino Gemelli IRCCS. Written informed consent was obtained from all patients.
After signing the informed consent, magnetic seed was placed into the center of the lesion under local anesthesia and ultrasound or stereotactic-guidance depending on the type of target [ Fig. 1]. The seed can be deployed by an 18 G preloaded needle of different length according to different breast sizes. Following insertion, mammograms in double projection (mediolateral oblique and craniocaudal views) are acquired to con rm correct positioning of the marker. The position of the seed was considered correct if positioned ≤10mm from the lesion. Two-view mammography was performed on the day of surgery to con rm the positioning of the seed and to exclude any migration. Intraoperative the Magseed is detected using the Sentimag probe.

Skin tattoo
Skin tattoo is conducted by use of mammography or sonography the same day or the day before the surgery. Ultrasound-guidance is reserved to lesions detectable by the ultrasound examination, performed by positioning the probe on the target, without applying pressure, with the patient supine and the ipsilateral arm abducted, in the surgical position [ Fig. 2]. The depth of the lesion under the skin was measured with ultrasound using minimal breast compression.
Stereotactic guidance is reserved in case of microcalci cations, architectural distortion and clip deployed at the end of biopsy, not visible by ultrasound. A metallic marker is put on the projection of the tumor on the skin surface based on previous mammography [ Fig. 2]. In case of big lesions or widespread microcalci cations, more than one marker is used to delimit the extension of the area. Two viewsmammogram (mediolateral oblique and cranio-caudal views) are performed to con rm the correct position of the metallic marker before performing of the skin tattoo.

Surgical specimen
Intraoperative specimen radiography was acquired during surgery to con rm the correct surgical excision. After that, the surgical specimen was sent to the pathologist for histological assessment.
All radiographs were obtained in two orthogonal planes, cranio-caudal and mediolateral oblique views.
Specimen radiographs were compared with baseline mammograms to assess adequacy of the excision in terms of presence both of the lesion and clip, and close margins. For this purpose, to make a more standardized assessment, we divided ideally the surgical specimen in three concentric circles [ Fig. 3].

Statistical analysis
All data and statistical analysis were carried out using SPSS (version 26.0, SPSS Inc., Chicago, IL, USA). Continuous variables were presented as means ± standard deviation (medians and interquartile ranges (IQR), whilst categorical variables were summarized as numbers and percentages. Fisher's exact test was used to compare categorical variables and Anova Test for continuous variables. A p < 0.05 was considered statistically signi cant. We compared demographics, breast size, site and size of the lesion, radiological and surgical characteristics between the two groups in order to highlight any differences.

Results
We retrospectively analyzed 77 patients from November 2020 to November 2021. Median age was 58.06 ± 12.53 (57; 48.5-68) years. Anatomic, epidemiological and radiological characteristics of enrolled patients are reported in Table 1. Thirty-seven magnetic seeds were placed in 36 patients (48.6%) and 40 skin tattoos were performed in the remained 40 patients (51.4%). There wasn't statistically signi cant difference for the histological diagnoses at the preoperative biopsy between the patients localized with magnetic seed and with skin tattoo (p = 0.976, Table 1). Also, no difference was found within ultrasoung ndings between the two groups (p = 0.280).
Concerning the breast size, in the skin tattoo group the highest percentage was small A-cup (47.5%), while in the seed grup was medium B-cup (62.2%), with a signi cant difference (p = 0.032, Table 1). The site of the lesions was similar in the study population (p = 0.536), with a prevalence of the upper-outer quadrant.
The depths of lesions in the breasts were statistically signi cantly different between the two groups (p = 0.009), with a higher prevalence of lesions deeper in the Magseed group (mean depth was 15.7 mm for magseed and 12.2 mm for skin-tattoo group -p < 0.0001), not deeper than 40mm (Table 1).
Thirty-three magnetic seeds (89.2%) were placed under ultrasound guidance with 4 (10.8%) being placed stereotactically. Twenty-three skin tattoos (57.5%) were performed under stereotactic guidance, 17 (42.5%) were ultrasound-guided [ Table 2]. The seeds were located correctly at the two-views mammogram acquired following the insertion in 97.6% (36/37) of cases [ Table 2]. No complications regarding placement of the magnetic seeds were reported. The time registered in relation to localization procedure and surgery date was longer in Magseed group (4.08 ± 15.8 days vs 0.40 ± 0.55 days, p = 0.146, Table 3). While time spent during localization was the same between the two groups (11.7 for skin-tattoo vs 13.2 minutes for magseed -p = 0.236). The time taken to detect the lesions after the skin incision was signi cantly different in the groups (p = 0.008), longer in Magseed cases (≥ 70 minutes in 12 cases localized with magseed and 2 cases with skin tattoo, Table 3). The identi cation and selective excision of the lesions marked was successful in 100% of patients. Magseed identi cation rate with Sentimag detector was 100% (37/37).
Magnetic seed was placed with more frequence in patients with small lesions (p = 0.017), smaller than 5mm in 18.9% of cases [ Table 3].
Regarding the ratio of breast volume to surgical volume, a statistically signi cant difference was observed in the seed group (0.98 ± 0.56 for Magseed vs 2.56 ± 3.05 for skin tattoo, p = 0.003), with less breast parenchyma removed. Stratifying the data according to lesions dimensions, magnetic seed removed a smaller area of breast parenchyma, especially in lesions smaller than 8 mm (0.93 ± 0.47 for Magseed vs 2.4 ± 2.1 for skin tattoo, p = 0.019, Table 3).
The accuracy of the resection assessed with surgical specimen was excellent in both groups, with 51.9% of the total lesions sited in circle 1, of whom 52.5% had undergone a skin tattoo localization, and 51.4% had undergone a Magseed insertion. No statistically signi cant difference was found in the groups regarding the distance between the lesion and the nearest margin (p = 0.518) of the surgical specimen at the radiography, as illustrated in Table 3.
Moreover, in most of the cases (64.9% of the total), an intraoperative widening was not necessary, performed in 40% of localization with skin tattoo and 29.7% with magnetic seed [ Table 3].
The re-excision rate for positive margins found in the cohort was 0%.

Discussion
Magseed (Endomagnetics, Cambridge, U.K.) uses a non-radioactive technology that overcomes the logistical problems and risks associated with using radioactive seed [9]. use, logistical advantages and re-excision rate compared to WGL [13][14].
Both studies concluded that Magseed is an effective and accurate alternative of preoperative breast lesion localization.
Our department surgically treats about 1400 breast cancer patients per year and the most used method for localization of non-palpable breast lesions is skin tattoo. In our experience this method is valid and effective, with a low rate of re-excision [15].
To the author's knowledge, this is the rst study comparing magnetic seed localization with skin tattoo technique.
Patients undergoing neoadjuvant chemotherapy were excluded from the study because at the time of this study Magseed was only licensed to be inserted for up to 30 days prior to surgery. Now the insertion is unlimited, and often performed at the time of biopsy.
We placed 37 seeds in 36 patients. In one case two seeds were located in the same beast, not in close proximity (< 2 cm), so it was easy to separate them. The magnetic seed was successfully located less than 10 mm from the lesion at the control mammogram in 97.6% of cases, maybe because of the good experience of our team in performing interventional procedures.
The time demanded in the localization with seed and skin tattoo was almost the same (12.6 minutes), even in this case, because of the familiarity of our breast radiologists with imaging-guided breast biopsy and localization procedures.
A correlation between localization technique, breast size and lesion depth was observed. The majority of magnetic seeds were placed in medium breasts (B-cup, 62.2%), compared with skin tattoo with a prevalence of small breasts (A-cup, 47.5%), and lesions located at a depth greater for the rst one (mean depth was 15.7 mm for magseed and 12.2 mm for skin-tattoo group -p < 0.0001). The seed was not placed deeper than 40mm from the skin surface, according to the detectable of the magnetic seed at greater depths [16].
Magseed was used with greater frequency in patients with small lesions (p = 0.017), smaller than 5 mm in 18.9% of cases.
This data may be used in the clinic to plan localization especially in case of small lesions in large breasts, because of the ease of detection of the seed with the Sentimag probe compared to skin-tattoo.
The longer time between localization and surgery in Magseed group (4.08 ± 15.84, p = 0.146) is due to the placement of the seed days before surgery, differently from the skin tattoo performed the day before or the same day of surgical procedure. It could reduce the weight of the breast radiology services and assist greatly with scheduling.
The migration of the seed from the initial placement was not formally analyzed with an objective parameter, but in our series no seed displacement was observed, and it was in line with the literature [14].
With both methods, 100% of the index lesions were completely removed and found in the surgical specimen. Two cases localized with the Magseed reported presence of DCIS in excision of the surgical margin, as the surgical treatment can be considered as oncologically successful.
A signi cant difference was observed in the overall surgical time (p = 0.008). The removal of magnetic seed required longer surgical time, despite the presumed advantage of Sentimag as an accurate tool in detecting target lesions. The reason is probably due to the surgeon's poor con dence with this new technique, compared with skin tattoo.
In our experience, Magseed does not differ in terms of oncological e cacy compared to skin tattoo. The site of the lesion in the surgical radiogram, the closest margin and the intraoperative widening were comparable in the two groups [ Table 2], according to the previous literature [17].
The less breast parenchyma removed in Magseed group in voluminous breasts (breast/surgical specimen volume ratio 0.98 ± 0.56 vs 2.56 ± 3.05, p = 0.003, Table 3), especially in case of small lesions, revealed that the seed allows to achieve a better aesthetic result, with a clinical value in terms of conservative surgery, not reducing the accuracy.
The re-excision rate reported was 0% for both groups, lower than the generally accepted percentage of 20-25%, (18) maybe due to the small sample size and the administration of loco-regional radiation therapy, according ASCO and ASRO [19].
Micha et al reported that Magseed is better tollereted than the pre-existing standard of guide wires. (20).
In our study we have not investigated this aspect.
Moreover, a point to consider is the cost of the Magseed, higher than the standard wire, and obviously than the skin tattoo.
Our study has several limitations. This was a single-center retrospective study, and Magseed was performed at the request of the multidisciplinary team, without a randomization. Despite the homogeneity of the two groups for the histological diagnosis, there is not a perfect match. Moreover, data on cost effectiveness were not evaluated. Due to the nature of the retrospective study, no data were collected on patient satisfaction of magnetic seed vs skin tattoo. Finally, the sample size was small. Our results need to be con rmed by prospective and larger studies.