SPECT/CT Lymphoscintigraphy Accurately Localizes Clipped and Sentinel Nodes After Neoadjuvant Chemotherapy in Node-Positive Breast Cancer

Abstract

Purpose 

The aim of this study was to evaluate the impact of SPECT/CT lymphoscintigraphy on targeted axillary dissection (TAD) in node-positive breast cancer (BC) patients who had undergone neoadjuvant chemotherapy (NAC).

Methods 

Sixty-two female BC patients with biopsy-confirmed axillary nodal metastases underwent NAC, followed by breast surgery with TAD. A metallic clip was placed in the sampled LN before NAC. On the day of surgery, a periareolar intradermal ^99mTc-nanocolloid injection was administered, followed by SPECT/CT lymphoscintigraphy. The clipped nodes were localized on CT images, assessed for ^99mTc uptake before surgery, and confirmed during the procedure.

Results 

T1–4, N1–2 patients were enrolled in the study. All patients underwent sentinel lymph node (SLN) biopsy. The clipped node was the SLN in 54 (88.5%) patients. In 3 patients (4.9%), a clip was found in a nonsentinel lymph node. In 4 patients, the clips were not visible on SPECT/CT images, and lymph nodes were not found during the procedure. SPECT/CT correctly localized the clipped lymph node in all patients. The overall false-negative rate for TAD was 3.33%. The mean follow-up duration was 29 months, and there were no axillary recurrences.

Conclusions 

SPECT/CT lymphoscintigraphy can accurately localize clipped nodes and SLNs after NAC in patients with node-positive BC.

Axillary nodal status remains a major prognostic factor for breast cancer (BC). In clinically node-positive (cN⁺) BC patients, Pilewskie and Morrow¹ reported nodal complete response in 35% to 63% of patients after neoadjuvant chemotherapy (NAC). Axillary lymph node dissection (ALND), which is the standard surgical approach for node-positive disease, has significant morbidity with unclear benefits in patients with a complete response.² The escalation of axillary surgery after NAC has accelerated since the SENTINA³ and ACOSOG Z1071⁴ trials reported high accuracy of sentinel lymph node biopsy (SLNB) after NAC. The main concern with SLNB after NAC is the false-negative rate (FNR) because of the possibility of blocked lymphatic drainage due to the fibrosis of channels or tumor emboli.^3,4 Dual-tracer mapping and removal of at least 3 nodes are recommended to overcome this problem. Targeted axillary dissection (TAD), which includes clip placement in the metastatic lymph node (LN) at initial diagnosis and performing SLNB with clip removal after completion of NAC, further decreases the FNR. In 2016, Caudle et al⁵ reported that TAD reduced the FNR from 10.1% to 4.2% when clipped nodes were removed and 1.4% when sentinel lymph node (SLN) and clipped LN were removed.

In 9% to 24% of cases, SLNs are not clipped nodes, and finding the clip may be a challenge during surgery.⁶ Clips are difficult to localize in the axilla using ultrasound (US), and placing a wire in the axilla is difficult and uncomfortable for patients. There is no defined perfect method for the TAD.

SPECT/CT is used for SLN identification in various diseases such as melanoma and head and neck cancers. This hybrid modality provides information about the anatomical location of SLNs with radiocolloid.

In this study, we assessed the guidance of SPECT/CT lymphoscintigraphy in TAD after NAC in cN⁺ BC patients. The primary end point was intraoperative detection of the clipped node. The secondary end point was FNR.

PATIENTS AND METHODS

Patients

This prospective cohort study examined SPECT/CT lymphoscintigraphy-guided TAD in women with cN⁺ BC who had undergone NAC. This study was approved by the Koc University Institutional Review Board (2021.145.IRB1.049).

Sixty-two women with locally advanced BC diagnosed between September 2017 and June 2022 who had biopsy-confirmed axillary nodal metastases were included in this study (Fig. 1). All patients underwent NAC, followed by breast surgery for TAD. Patients with recurrent, inflammatory, or metastatic BC; previous axillary surgery; radiation therapy; or pregnancy were excluded. Data were collected from the medical records, including operative notes, radiology, and pathology reports.

FIGURE 1: Study flowchart.

Breast cancer was diagnosed by core biopsy. The histopathologic type, hormone receptor, and human epidermal growth factor receptor 2 (HER2) status were determined. All patients underwent mammography (MG), breast US, breast MRI, and ¹⁸F-FDG PET/CT scans and were staged according to the eighth edition of the American Joint Committee on Cancer staging manual.

Breast Imaging and Clip Placement

Patients at our institution routinely undergo breast and nodal ultrasonography at diagnosis, midchemotherapy, and before surgery. Response to chemotherapy was also assessed using pre- and post-NAC breast MRI.

The most suspicious axillary LNs on US were biopsied using core-needle or fine-needle aspiration. If metastasis was identified, a metallic clip (Ultra Clip II Tissue Marker 17-gauge × 12-cm Needle-Ribbon; BARD, Inc, New Providence, NJ) was placed in the index-sampled LN before NAC. This procedure was performed by the same breast radiologist under US guidance. The index node was chosen based on the following criteria: loss of the fatty hilum, asymmetric cortical thickening (>3 mm), abnormal LN shape (round form), increased peripheral blood flow, and irregular cortical borders.

Neoadjuvant Treatment

Anthracycline and/or taxane-based chemotherapy regimens with the addition of HER2-targeted therapy were used if HER2 was positive. The medical oncologist chose the regimen.

SPECT/CT and Lymphoscintigraphy

On the day of surgery, all the patients underwent lymphoscintigraphy. Periareolar 0.5 mCi ^99mTc-nanocolloid was administered intradermally, followed by dynamic, planar, and SPECT/CT imaging using a hybrid system (Optima NM/CT 640; GE Healthcare). SPECT data were acquired for the thoracic region (matrix size, 128 × 128 pixels; 6° angle steps, 30 s/frame). The CT acquisition parameters were as follows: 140 keV, pitch 1.25, rotation time 1 second, and slice thickness 2.5 mm. The skin projection of the SLN was marked under the guidance of a gamma probe and SPECT/CT images. The images were interpreted at the workstation by a nuclear medicine specialist who assisted the surgeon (Fig. 2). The clipped LN and SLNs were localized on SPECT/CT images and detected whether they had ^99mTc uptake before surgery, which would then be confirmed intraoperatively.

FIGURE 2: A 40-year-old woman with HER2-positive locally advanced left BC. A, Initial PET/CT scan with intense FDG uptake in a metastatic axillary LN. B, Complete metabolic response in clipped LN after NAC. C, SPECT/CT lymphoscintigraphy (^99mTc-nanocolloid) showing the clips in the SLN. D, Clip placement under US before NAC. E, Clipped SLN seen on specimen radiography. F, Example of a clip seen on cut surface of SLN.

Surgical Management

The operations were performed by 2 breast surgeons. The type of surgery was determined by the surgeon as breast-conserving surgery, simple mastectomy, or nipple/skin-sparing mastectomy with reconstruction and SLNB with or without ALND.

In addition to the ^99mTc-nanocolloid injection, 18 patients also received 5 mL of blue dye (isosulfan blue) administered to the subareolar area according to the surgeon's preference. A gamma probe (Crystal Photonics SG04, Germany) was used to localize the SLNs. When the activity count on the LN was 10 times higher than the background, it was considered SLN and dissected. Blue-stained LNs and additional suspicious palpable LNs were removed as SLNs or nonsentinel lymph node (nonSLNs). Clip removal was confirmed by specimen radiography or macroscopic evaluation. All the dissected SLNs were sent to frozen sections. Axillary lymph node dissection was our standard, if the clipped LN was not found or the SLN was metastatic.

Pathologic Evaluation

Lymph nodes sent for frozen section analysis were evaluated by a breast pathologist. The presence of clips and/or blue dye was noted after macroscopic evaluation. Each LN was sliced along its long axis at a thickness of 2 mm. The shape, imprint, and/or frozen section slides of each LN were examined. All slices were subjected to permanent examination. At least 4 sections were obtained from each block of a sentinel node at 50-μm intervals and stained with hematoxylin-eosin. During sectioning, 1 spare section was used for cytokeratin immunohistochemistry. Sentinel lymph nodes containing tumor cells detected by hematoxylin-eosin or cytokeratin staining were considered positive according to the eighth edition of the American Joint Committee on Cancer. Pathological findings related to chemotherapy response were also recorded, including fibrosis and foamy macrophages.

Statistical Analysis

A false-negative event was defined as when an SLN or the clipped node was not metastatic, but there was metastasis in the residual axillary LNs. SPSS version 28.0 software was used for statistical analysis to calculate the median and range values.

RESULTS

Sixty-two patients were enrolled in the study. The median age of the patients was 46 years (range, 25–74 years). None of our patients had distant metastases based on ¹⁸F-FDG PET/CT findings. Ten (16.3%) of 61 patients had T1, 40 (65.5%) had T2, 9 (14.7%) had T3, and 2 (3.2%) patients had T4 tumors. All patients had ipsilateral axillary LN metastasis. Fifty patients (81.9%) had N1 clinical stage at presentation, and 11 patients (18.1%) had N2.

Only 1 clip was placed in the dominant LN. All patients received NAC, including 4 cycles of adriamycin-cyclophosphamide followed by taxanes. Trastuzumab, with or without pertuzumab, was added to taxane therapy in 10 HER2-positive BC patients (16.3%). After NAC, 9 patients (14.7%) underwent mastectomy, 15 (24.5%) underwent breast-conserving surgery, and 37 (60.6%) underwent nipple-sparing or skin-sparing mastectomy. The patient characteristics are shown in Table 1.

TABLE 1 - Clinicopathological Characteristics of the Patients

	All Patients (n = 61), n (%)
Age, median, y	46 (25–74)
Clinical T stage at presentation
cT1	10 (16.3)
cT2	40 (65.5)
cT3	9 (14.7)
cT4	2 (3.2)
Clinical N stage at presentation
cN1	50 (81.9)
cN2	11 (18)
Type of surgery
Mastectomy	9 (14.7)
BCS	15 (24.5)
NSM/SSM	37 (60.6)
No. sentinel nodes removed, mean	2.5
Axillary dissection	34 (55.7)
cN1	28 (45.9)
cN2	6 (9.8)
Pathological response in the axilla
Complete	40 (65.5)
Partial	21 (34.4)
Histology
Ductal	47 (77)
Lobular or mixed	14 (22.9)
Subtype
HR+/HER2−	42 (68.8)
HER2+	10 (16.3)
HR−/HER2−	9 (14.7)

BCS, breast-conserving surgery; HR, hormone receptor; NSM, nipple-sparing mastectomy; SSM, skin-sparing mastectomy.

Targeted axillary dissection was performed in all patients. In 1 patient, SLN mapping was unsuccessful, possibly due to the blockage of lymphatic vessels. That patient was excluded from the study. The mean number of SLNs was 2.5 (range, 1–6). Thirty-four patients (55.7%) had confirmed metastases intraoperatively and underwent further ALND, whereas 27 (44.3%) underwent SLNB only. Blue dye was used in 18 patients (29.5%), all of whom were hot nodes. Blue dye did not identify an additional SLN in any of the patients. Forty patients (65.5%) had a complete response in the axillary LNs, whereas 21 (34.4%) had a partial response.

SPECT/CT identified 1 to 5 (mean, 1.8) SLNs in each patient. In 54 patients (88.5%), the clipped node was the sentinel node, which was confirmed by surgery. In 3 patients (4.9%), the clip was found in the nonSLN on both SPECT/CT and surgery (Fig. 3). One of these patients had guidewire localization of the clipped node before removal. The other 2 patients underwent ALND, as they had metastasis to the SLNs. In the remaining 4 patients (6.4%), the clips were not localized by SPECT/CT and could not be found during surgery, despite radiographic screening. Those patients had ALND due to metastases in the SLNs; however, the clips were not detected on specimen MG. Only 1 of the 4 patients had an identified migrated clip seen on follow-up MG. In all patients (100%), SPECT/CT correctly localized the clipped LN as the SLN or nonSLN (Table 2).

FIGURE 3: A 38-year-old woman with triple-negative locally advanced left BC. A, Intense FDG uptake of the metastatic LN. B, Complete metabolic response in clipped LN after NAC. C, SPECT/CT lymphoscintigraphy (^99mTc-nanocolloid) identified the clipped LN as nonSLN. D, SPECT/CT lymphoscintigraphy (^99mTc-nanocolloid) showing the SLN in another section.

TABLE 2 - Accuracy of SPECT/CT localization

	SPECT/CT, n	Intraoperative, n
Clips in the SLNs	54	54
Clips in the nonSLNs	3	3
Clips not identified	4	4
Accuracy	100%

The FNR was calculated for the 34 patients who had undergone ALND. One patient had a negative SLN, but metastasis in a nonSLN. In this subgroup, the FNR was 2.9%.

The patients were followed up at 6-month intervals by physical examination, US, and/or MG. There was no axillary relapse at the median follow-up of 29 months (8–65 months).

DISCUSSION

This study shows that SPECT/CT lymphoscintigraphy may be a valuable tool and a facilitator for localizing the clip in the axilla and the SLN before surgery. There is a great variety between institutions and countries in the techniques used to remove the targeted LN in initially cN⁺ patients who downstage to cN⁻ after NAC.^7–9 Metallic clip placement, MARI (marking the axillary lymph node with radioactive iodine seeds) procedure,¹⁰ tattooing,¹¹ and the Savi Scout reflector¹² are the major techniques used for localization. Radiation safety protocols have limited the widespread use of radioactive seeds. However, it may be challenging to identify the clip during surgery; clips may migrate out of the LN during chemotherapy. Preoperative localization of the clipped LN has been reported to be unsuccessful in 5% to 30% of patients.¹³ In our series, 85% of sentinel nodes contained clips. SPECT/CT images correctly localized the clips in the SLN, as confirmed during surgery. Caudle et al⁵ reported that in 23% of cases, the clips were not in the SLN. Kuemmel et al¹⁴ recently published a multicenter registry study showing that in 64.8% of patients, the SLN and clipped node were identical. In approximately 65% to 85% of cases, the SLN will contain the clips, and any attempt at preoperative localization will be unnecessary in those patients. Routine preoperative wire localization for every patient is questionable and should not be recommended. SPECT/CT can detect patients whose SLN and clipped nodes are not identical and can be candidates for preoperative localization.

CT can detect clip markers because of their metallic content and has been used for localization.¹⁵ Lymphoscintigraphy and SLNB with intraoperative gamma probe are safety-proven axillary staging techniques in BC.¹⁶ Planar lymphoscintigraphy imaging identifies SLNs in more than 95% of patients with BC, with issues of increased cost and availability.¹⁷ With the fusion of SPECT and CT images, it is possible to localize the tumor anatomically and verify whether the clipped axillary LN is radiotracer avid. SPECT/CT has been shown to improve SLN identification in overweight patients; moreover, SLNs were identified in 75% of patients in whom the blue dye technique failed.¹⁸ Sentinel lymph node biopsy after NAC may be another area in which SPECT/CT is superior to planar lymphoscintigraphy.

It would be more accurate and easier for a surgeon to open the axilla knowing where the target is, rather than opening and looking for it. SPECT/CT shows anatomically where the SLN and the clips are, and with the handheld gamma probe, SLN can be precisely excised. When the clipped node is away from the SLN, it can be marked with a wire or another localizer, depending on the surgeon's choice. If the clipped node is not the SLN but lies nearby, the surgeon can still find it under the guidance of CT images and palpation, with intraoperative specimen MG verification. Barrio et al¹⁹ proposed that patients with cN1 disease rendered cN0 with NAC, with 3 or more negative SLNs with SLNB alone, and nodal recurrence rates were low, without routine clipping. This report is behind the discussion on leaving lost clips in place. In the future, we possibly will not fall after each lost or migrated clips in selected cases. In our study, when the sentinel node was not the clipped node and clips could not be identified by intraoperative fluoroscopy, ALND was performed in every case.

In 3 (4.5%) of our patients, the clipped LNs were nonSLNs, and SPECT/CT had correctly localized the LNs. In the first 2 cases, the SLNs were metastatic, and ALND was performed, proving clips in the nonSLNs. In the third case, as we learned from our previous experience, we placed a guidewire in the clipped node before surgery and removed it with the SLN. In the remaining 4 patients (6.5%), we could not locate the clips either with SPECT/CT or during surgery. All these patients had ALND; however, the clips were not observed on specimen mammographies. In only 1 patient's follow-up MG, migrated clips could be identified in the axillary tail. In the first cases of axillary clip placement and surgery, we also had a learning curve. We reported our preliminary results with the first 38 patients in which SPECT/CT images were 100% correlated with intraoperative findings.²⁰ These results are consistent with those of the current study with an increased number of patients. There was no axillary relapse at the median follow-up of 29 months (range, 8–65 months).

Newer wireless, nonradioactive localization devices are US Food and Drug Administration–approved for long-term insertion before surgical retrieval. Recently, Gallagher et al²¹ published 86 node-positive patients who had radar-localized reflectors directed toward TAD and completed ALND after NAC. They reported a high accuracy with a 5.1% FNR. Baker et al¹² reported the results of 23 patients who used the Savi Scout probe and the dual SLNB method for TAD after NAC. The Savi node was the SLN in 87% of cases. To date, no method has been found to be perfect, but a detector probe may be wiser to use than a silent metallic clip if one cannot show where it is. The cost-effectiveness of the procedures also needs to be evaluated and compared.

Has Simsek et al⁹ have reported the results of SPECT/CT lymphoscintigraphy in 15 patients to determine clipped LNs during SLNB after NAC. In 20% of patients, clips were detected in the nonSLNs, and wire localization was performed to remove the clipped LNs with the SLNs. This limited study, similar to our 61-patient cohort, reports the technique as feasible and easy to adapt.

CONCLUSIONS

Our results show that SPECT/CT ^99mTc-nanocolloid lymphoscintigraphy can accurately localize the clipped and SLNs after NAC in initially cN⁺ BCs. SPECT/CT facilitates detection of the clipped LN. In addition to being comfortable for patients, it is also convenient for the surgeons.