Savi Scout® wireless localisation of breast and axillary lesions: lessons learned from Singapore’s early experience

Abstract

INTRODUCTION Hookwires are commonly deployed under imaging guidance to localise non-palpable breast lesions for excision. However, the use of hookwires has some disadvantages, including patient discomfort, wire migration, damage to surrounding anatomical structures, surgery scheduling inconveniences and limited access to axillary nodes. Novel, alternative, non-radioactive wireless localisation devices using technologies such as radiofrequency identification, magnetic seed and radar have been developed to address these shortcomings.[1,2] While these devices have seen increasing usage in America and Europe, they were introduced to Asia only recently. One of these wireless techniques, the Savi Scout® (SS) surgical guidance system (Cianna Medical, Merit Medical Systems, Inc., South Jordan, UT, USA), was made available in Asia in 2019, and Singapore was the first Asian country to utilise it for breast and axillary localisations. Savi Scout employs radar technology, and it received the United States Food and Drug Administration clearance in 2014. The SS consists of a reflector implant, a needle introducer and an external check console. The reflector is a 12-mm metallic implant [Figure 1] consisting of thin nitinol antennae protruding from either end of a central transistor body. It is inserted percutaneously into the soft tissue via a single-use, preloaded, 16-gauge needle introducer and is deployed by uncovering the overlying sheath at the distal end of the introducer. This passive reflector delivery mechanism prevents damage to the thin antennae. The introducer is unsheathed by first unlocking the release button to either left or right and then retracting it along a sliding track [Figure 1]. Once the reflector is deployed, it cannot be repositioned. A handheld probe connected to the check console is used to locate the deployed reflector by transmitting a radio wave signal (radar), which is received and reflected back by the reflector. The signal capture and reflection mechanism of the reflector is multidirectional and is used to guide direction and distance to the target up to a depth of 6 cm. Unlike hookwires, SS does not have any components protruding from the skin. In addition, there is no placement expiry after deployment and it can be deployed at any time before surgery day, which provides flexibility in procedural scheduling.Figure 1: Photograph shows the parts and functions of the Savi Scout® needle introducer system and the 12-mm-long reflector (inset).Studies from America and Europe have evaluated SS to be a safe and convenient localisation technique for the breast and axilla.[3-8] However, to the best of our knowledge, there have been no published reports evaluating its performance in Asian women, and it is unclear if dense breast tissue, which is more prevalent in Asian women, may affect SS deployment and signal detection. We described our experience in the initial use of SS in Singapore women with the aims of providing an assessment on its performance in Asian women and finding ways to optimise its use. METHODS This was an institutional, review board-approved retrospective review of patients who underwent imaging-guided SS localisations at multiple centres in Singapore from July 2019 to June 2021. Performance was evaluated by 13 users, six of whom were breast imaging-dedicated radiologists and seven were breast surgeons. The ease of reflector deployment, time taken to deploy the reflector, postdeployment signal detection, incidence of reflector damage, incidence of reflector malpositioning, incidence of reflector migration, radiological visibility of the reflector, ease of intraoperative localisation, surgical retrieval rate, complication rate and overall user satisfaction were evaluated. Malpositioning was defined as the centre of reflector sited more than 10 mm from the epicentre of the lesion at deployment. Reflector migration referred to displacement of more than 10 mm from its original deployment site. Deployment time was the duration interval between introducer entry and exit of the skin. Assessments were graded on a 4-point scale of none, low/mild, moderate and high/severe categories, where appropriate. Statistical tests were performed with the online software GraphPad QuickCalcs (https://www.graphpad.com/quickcalcs/). Continuous variables were compared using the Student’s t-test. Comparisons of categorical variables were performed with Fisher’s exact test. The differences were considered statistically significant at P < 0.05. RESULTS Forty-four reflectors were deployed for 31 breast lesions and 13 axillary nodes in 40 female patients. Planned deployments had to be cancelled for five other women with a history of nickel allergy. Thirty-eight (86.4%) reflectors were deployed under ultrasound guidance, while six (13.6%) were performed under mammogram guidance. Four (9.1%) breast placements were inserted before commencement of neoadjuvant chemotherapy, with a mean placement duration of 184 days. There were ten (22.7%) placements in metastatic nodes that were inserted before neoadjuvant chemotherapy for subsequent targeted axillary dissection (TAD), with a mean placement period of 148 days. Twenty-one (47.7%) reflectors were inserted on the same day as surgery and another nine (20.5%) were non-same day deployments within 2 weeks of surgery. All users expressed overall satisfaction with SS use and there were no major complications. On specific technical aspects of deployment, there was no significant resistance when advancing the introducer in dense breast tissue. The introducer’s release button was highlighted to be mechanically stiff in 12 (27.3%) instances, with four (9.1%) encountering mild resistance and eight (18.2%) encountering significant resistance. Mean deployment times were 3 min and 9 s and 1 min and 53 s (P = 0.037) for insertions performed under ultrasound and mammogram guidance, respectively. There were no cases of reflector damage, malpositioning or migration [Tables 1 and 2].Table 1: Deployment outcomes by lesion and deployment types.Table 2: Surgical technical performance of deployments (N=44).On the surgical technical aspect, five of seven breast surgeons (71.4%) felt that SS provided greater flexibility in incision placement. Intraoperatively, the reflector was easily and accurately located with the use of the check console, apart from a few cases that encountered signal detection difficulty. Specimen radiographs confirmed that the reflectors were all fully retrieved with no evidence of damage. Failed signal detection on immediate postdeployment was encountered in three (6.8%) cases. In two of these cases, the reflector was inserted into a metastatic axillary node before commencement of neoadjuvant chemotherapy for the purpose of TAD. In both instances, core needle biopsy of the nodes was performed just before reflector deployment and postbiopsy haematomas were the likely cause of signal impedance. Signal was detected for both reflectors during a signal check a week later. At surgery 5 months later, signal was detected intraoperatively for one of the cases, but was absent in the other (2.3%). In this case of the absent intraoperative signal, the targeted node and the reflector were successfully retrieved after surgical exploration with the aid of intraoperative ultrasound. Subsequent investigations showed no malfunction of the reflector and check console, with no definite cause found for the signal failure. The third case of signal failure was related to a mammogram-guided deployment on surgery day, targeting a postbiopsy clip. There was a residual moderate-sized, postbiopsy haematoma next to the clip, which was likely impeding the signal. The clip was subsequently localised with a hookwire as an additional measure, but there was intermittent signal detected intraoperatively, which was sufficient to guide excision without difficulty. There was also one (2.3%) case of intermittent signal in a case of ultrasound-guided deployment adjacent to a moderate-sized, post-core needle biopsy haematoma. There was mild difficulty locating the reflector intraoperatively, but excision was successful. Signal loss triggered by diathermy contact was encountered in two cases (4.5%) that involved wide local excision for breast cancer. In both instances, the reflector tip was visible in the dissection plane at the time of signal loss, and the reflectors and tumours were excised with clear margins. The reflector was well visualised on ultrasound in 30 out of 31 cases (96.8%) and poorly visualised in one case (3.2%). The reflector was distinctively visualised on all postdeployment mammograms that were performed (30/30) [Figure 2]. In the six cases that had breast magnetic resonance imaging (MRI), the reflector was well seen in all of them and was best visualised in the T1-weighted, non-subtracted, postcontrast sequences with fat saturation. The reflector was seen as a small blooming artefact on MRI.Figure 2: Radiological appearance of reflectors (arrows) in a patient. (a) Mammogram shows the reflectors inserted before neoadjuvant chemotherapy embedded within the breast tumour and metastatic axillary node. (b) Sonogram shows the echogenic linear reflector within the enlarged node.DISCUSSION Our multicentre early experience demonstrated the utility of SS in a variety of clinical settings with good outcomes in Asian women. There was no signal inhibition in dense breast tissue and the introducer was able to advance into dense breasts without difficulty. Unlike hookwires, the small reflector posed little risk of injuring adjacent structures and was a useful alternative to hookwires for difficult-to-access sites such as lesions close to the chest wall or axillary nodes close to vital anatomical structures. Flexibility in scheduling deployments was a welcomed benefit. The implant also had the advantage of causing minimal MRI artefacts, which did not significantly obscure MRI assessment. There were, however, a few drawbacks encountered. Rare instances of intermittent or absent reflector signal were documented, mostly related to signal impedance by an adjacent haematoma, which is well documented as a cause of signal failure.[6,7] To reduce the risk of signal failure, the reflector is best placed superficial to the haematoma if present. True reflector malfunction is extremely rare, but it has been documented before.[3,6,8] We recommend performing a signal check just before surgery, so that alternative means of localisation can still be arranged before surgery in the rare event of signal failure. Another intraoperative pitfall was reflector deactivation after electrocautery contact. To avoid this, there is a need to carry out dissections with greater care when nearing the reflector and use the check console to guide proximity distance from the reflector frequently. Even with reflector deactivation, we did not feel it was a significant problem because it would indicate that the targeted site has been reached and the reflector together with the lesion will be identified in the dissection plane by then. In our series, the two cases of reflector deactivation did not adversely affect surgical outcomes. In the event that the reflector cannot be detected and located during surgery, it can be localised with the aid of intraoperative imaging. Being easily seen on sonography and X-ray examinations, the reflector may be located with the help of intraoperative ultrasound or fluoroscopy, as in one of our cases. The introducer’s delivery mechanism was slightly cumbersome. The needle tip had to be advanced a further 6 mm, so that the reflector would be centred well upon deployment. For ultrasound deployments, the proceduralist may take a few adjustments to achieve this optimal needle position, whereas this additional advancement can be precalibrated into the mammogram machine settings, removing the need for manual adjustments in mammogram-guided localisations. This was reflected in the longer procedural times for ultrasound-guided deployments compared to mammogram-guided ones. Notably, there was unexpected difficulty in retracting the introducer’s release button during reflector deployment. A few cases required extreme effort to fully retract the button and the procedure took longer to complete. It was not an isolated batch problem and there were also no such reports in other studies for reference. To troubleshoot this, our users found it easier to retract after switching the button to the opposite sliding track. The needle position may also unintentionally shift during the physical struggle to retract the release button, especially when performing under ultrasound guidance. We suggest checking the needle position when the release button is retracted halfway. If there is a need to reposition, the release button may be reversed to resheath and protect the reflector before proceeding to reposition the needle. We found that the reflector was not damaged with this manoeuvre. On the surgical technical aspect, most of the breast surgeons felt that SS allowed greater flexibility in incision placement. The accurate localisation facilitated a more targeted resection, which can help reduce excessive tissue resection, and skin incisions could be confidently placed directly over the lesion or cosmetically placed far from the lesion. Finally, SS use was limited by nickel allergy and cost considerations. The nitinol antennae contain small amount of nickel and, out of caution, we did not proceed with SS insertion for a few women who had nickel allergy.[4,5] Nickel allergy, however, is fairly prevalent and clinicians should routinely ask patients about this, so that they can plan in advance for alternative localisation methods, if necessary. Cost-wise locally, the reflector can be up to nine times more expensive than a hookwire. A significant capital outlay would also be necessary for procurement of the console. Unlike the findings from several American and European studies, many local clinicians felt that SS was too expensive for routine localisations despite its perceived benefits, which largely explained the low usage during the study period. However, we found that the most cost-effective utility in our local setting was the upfront deployment of SS in patients undergoing neoadjuvant chemotherapy before a planned breast conservation surgery or TAD. This group of patients will fully benefit from the dual functionality of SS acting as both a tumour clip and a localisation guide for subsequent surgical excision, eliminating the need for another localisation on surgery day, which will save procedural costs. One of the limitations of this paper is not evaluating the patients’ experience. However, none of the patients explicitly expressed dissatisfaction or discomfort with the reflector deployed in them. Cosmetic outcomes were also not evaluated. These are important points that can be explored in future. In summary, SS worked well in Asian women with some clinical benefits, but users need to be aware of its limitations. We have suggested ways on how to troubleshoot and optimise its use, and we hope our early experience will help other institutions in Asia to implement SS or other wireless localisation devices in their clinical practice and to navigate potential challenges. Acknowledgement We would like to thank Dr Permeen Akhtar bt Mohamed Yusoff from Research Office, Singapore General Hospital, Singapore, for assistance in editing and formatting of the manuscript. Financial support and sponsorship Nil. Conflicts of interest There are no conflicts of interest.