Optimal efficient imaging time for identification of sentinel lymph node with Tc-99m Tilmanocept in breast cancer patients

Abstract

We determined the optimal imaging time for axillary lymph node (LN) visualization following Tc-99m Tilmanocept in breast cancer patients to establish imaging guidelines that can allow for a reliable and efficient yet high yield study prior to surgery. Retrospective analysis in 651 patients who underwent lymphoscintigraphy, comparing LN visualization on immediate, 15-minute, and 90-minute delayed imaging after injection of Tc-99m Tilmanocept. Statistical analysis was performed using McNemar’s test, kappa coefficient, and Pearson Chi-square test. Five hundred and six patients had either immediate or immediate and 90-minute delayed imaging. Of these patients, 203 (40.1%) had both immediate and 90-minute delayed images. Of these 203 patients, 54 (26.6%) had ≥1 lymph node(s) identified immediately and 196 (96.6%) had ≥1 lymph node(s) identified at 90 minutes (P<0.0001). A kappa coefficient of .0256 was observed (95% CI: .0058-.0453). One hundred and forty-five additional patients had 15-minute delayed imaging. Of these patients, 117 (80.7%) had ≥1 lymph node(s) identified, which was significantly fewer compared to the number of patients with ≥1 lymph node(s) detected at 90 minutes (P<0.0001). Ninety-minute delayed imaging is optimal for identifying sentinel lymph node(s) following Tc-99m Tilmanocept injection in breast cancer patients.

Introduction

Breast cancer is the most common cancer diagnosis in women in the United States and the leading cause of cancer death in women worldwide [1, http://gco.iarc.fr/today/data/factsheets/populations/900-world-fact-sheets.pdf]. Once breast cancer is diagnosed, staging provides guidance for appropriate management strategies as well as information regarding prognosis. Lymph node involvement is one of the main factors in determining breast cancer stage [2].

Sentinel lymph node (SLN) is the first or group of first lymph nodes in the path of lymphatic drainage from the site of malignancy. Current guidelines strongly recommend sentinel lymph node biopsy (SLNB) to evaluate axillary lymph nodes in individuals with early-stage breast cancer who do not have known nodal metastases and in individuals with prior breast and/or axillary surgery [3]. In addition, SLNB was found to decrease the need for more invasive axillary lymph node dissection in patients without evidence of metastatic disease [4,5].

Sentinel lymph nodes are often localized by radioisotope labeling. Lymphoscintigraphy can be performed prior to surgery to image radiolabeled nodes following radioisotope injection [6,7]. These images provide a roadmap of lymphatic drainage and axillary lymph nodes facilitating sentinel lymph node identification at the time of surgery [8-11]. A recent meta-analysis reported that the sentinel lymph node identification rate is 98.4% when lymph nodes are visualized on lymphoscintigraphy as compared to 84% when lymph nodes are not visualized on lymphoscintigraphy [12]. Therefore, lymphoscintigraphy and lymph node visualization prior to surgery may be essential.

Currently available radiotracers include Tc-99m sulfur colloid and Tc-99m Tilmanocept (trade name: Lymphoseek®). Tc-99m Tilmanocept, which binds mannose receptors (CD206) expressed within lymph nodes, has been shown to be a superior radiotracer to Tc-99m sulfur colloid due to improved SLN targeting and faster clearance from the injection site [13]. As a result, many institutions perform lymphoscintigraphy with Tc-99m Tilmanocept. While studies suggest that delayed imaging may be unnecessary for Tc-99m sulfur colloid, no studies to date have shown the optimal imaging time for lymphoscintigraphy following Tc-99m Tilmanocept injection in breast cancer patients [14]. Therefore, standard operating procedures are oftentimes based on data from Tc-99m sulfur colloid with imaging times following administration of Tc-99m Tilmanocept varying between institutions and between referring surgeons.

The objective of this study is to determine the optimal imaging time for axillary lymph node visualization following Tc-99m Tilmanocept injection. Understanding optimal imaging time is important in establishing imaging guidelines that can allow for an efficient yet high yield study prior to surgery.

Materials and methods

The Institutional Review Board of Duke University approved this Health Insurance Portability and Accountability Act compliant retrospective single-institution study and waived the requirement to obtain informed consent.

Patients

Review of our electronic medical record procedure worklist (Epic Hyperspace® v.May 2019, Verona, WI) identified 506 patients with breast cancer diagnoses who received lymphoscintigraphy procedures performed between February 1, 2016 and May 31, 2017 with immediate and 90 minute delayed imaging. In addition, 145 additional patients with breast cancer diagnoses who received lymphoscintigraphy procedures between Feb 1, 2018 and December 31, 2018 with 15 minute delayed imaging were studied. Bilateral lymphoscintigraphy injections were performed in 24 patients, who were diagnosed with bilateral breast cancer. There were no exclusion criteria.

Lymphoscintigraphy and imaging technique

Lymphoscintigraphy was performed by injecting 2 mCi (0.2 mL) for next day surgery or 0.5 mCi (0.1 mL) for same day surgery of Tc-99m Tilmanocept using a 1 mL syringe with a 26 gauge ½ inch needle. Various injection techniques were utilized based on the preference of the 5 referring breast surgeons: 1) intradermal injection of 2 mCi of Tc-99m Tilmanocept in the periareolar region of the upper outer quadrant of the affected breast; 2) intradermal injection of 1 mCi of Tc99m Tilmanocept and subcutaneous injection of the remaining 1 mCi in the vicinity of the known malignancy. Lymphoscintigraphy was performed by nuclear medicine physician assistants, nuclear medicine fellows, and 5 nuclear medicine fellowship trained radiologists.

Subsequently, per institutional standardized operating protocol, anterior/posterior, lateral, and oblique planar images were obtained either immediately only or immediately and at 90 minutes following radiotracer administration based on surgeon preference using several gamma cameras (Innova IG 630, Infinia Hawkeye, Infinia and Discovery 670, General Electric Healthcare, Chicago, IL). Immediate images were obtained on all 506 patients. If radiotracer activity was visualized in an axillary lymph node on the immediate image, no further imaging was performed for 4 of the 5 referring surgeons. If no axillary lymph node was visualized on the immediate imaging, 90 minute delayed imaging was performed for 4 of the 5 surgeons. For 1 of the 5 surgeons, both immediate and 90 minute delayed imaging were performed regardless of the results of the immediate imaging.

In the 145 additional patients who were studied, Tc-99m Tilmanocept injection was performed as described above and anterior/posterior, lateral, and oblique planar imaging were obtained at 15 minutes using several gamma cameras (Innova IG 630, Infinia Hawkeye, Infinia and Discovery 670, General Electric Healthcare, Chicago, IL). Lymphoscintigraphy imaging was interpreted by 5 nuclear medicine fellowship trained radiologists with 8-31 years of experience.

Data collection

We reviewed radiology reports on the Picture Archiving and Communication Systems (PACS) (Centricity PACS v.6.0, GE Healthcare, Chicago, IL) for each case. No new interpretations of the diagnostic imaging were performed for the purposes of this study. We reviewed radiology reports and obtained information regarding the date of procedure/imaging, name of referring surgeon, injection dose and technique, timing of images obtained following injection, and number of lymph nodes identified on imaging.

Data analysis

Timing of images was classified as immediate, 15-minute delayed, and 90-minute delayed. At each time point, cases were categorized as demonstrating either 0 or ≥1 lymph node(s). Statistical differences in LN identification between patients with both immediate and 90-minute delayed images were assessed using McNemar’s test. Additionally, the kappa coefficient was calculated to determine agreement between immediate versus 90-minute delayed imaging in these same patients. LN identification in cases with 90-minute delayed images (n=203) was compared to cases with 15-minute delayed images (n=145) using Pearson Chi-square testing.

Results

There were a total of 506 patients who had either immediate, 90-minute delayed, or both immediate and 90-minute delayed imaging (Figure 1). Of these 506 patients, 203 patients (40.1%) had both immediate and 90-minute delayed imaging. Of these 203 patients, 54 (26.6%) had ≥1 lymph node(s) identified with immediate imaging and 196 (96.6%) had ≥1 lymph node(s) identified at 90 minutes, which was statistically significant (P<0.0001). A kappa coefficient of .0256 was observed (95% CI: .0058-.0453), indicating poor agreement.

Figure 1.

Immediate versus 90-minute delayed images. A. Immediate image demonstrates radiotracer activity at the injection site (arrow head), radiotracer activity within the lymphatic channel (thin arrow), and radiotracer activity in a single axillary lymph node (thick arrow). B. Image obtained 90 minutes after Tc-99m Tilmanocept injection demonstrates multiple axillary lymph nodes. C. 90-minute delayed image using Co-57 transmission source better outlines the body.

In addition, 145 patients had 15-minute delayed imaging. Of these 145 patients with 15-minute delayed imaging, 117 (80.7%) had ≥1 lymph node(s) identified (Figure 2). The difference in identification of ≥1 lymph node(s) on 15-minute (80.7%) and 90 minute (96.6%) delayed imaging was significant (P<0.0001) (Table 1).

Figure 2.

Immediate versus 15-minute delayed images. A. Immediate image demonstrates radiotracer activity at the injection site (arrow head) and within the lymphatic channel (thin arrow). B. Image obtained 15 minutes after Tc-99m Tilmanocept injection demonstrates radiotracer activity at the injection site (arrow head), within the lymphatic channel (thin arrow), and in a single axillary lymph node (thick arrow). C. 15-minute delayed images using Co-57 transmission source.

Table 1.

Imaging results

	Immediate	15 min	90 min
Total (N)	203	145	203
#Patients with ≥1 LN* (N [%])	54/[26.6]	117/[80.7]	196/[96.6]

^*Lymph Node.

Discussion

Lymphoscintigraphy is widely utilized to visualize radiolabeled axillary lymph nodes prior to SLNB. To our knowledge, this is the first study to investigate the optimal imaging time following Tc-99m Tilamanocept injection in breast cancer patients. Our study included a large total number of 651 cases. There was a comparable distribution of cases with immediate, 15-minute delayed, and 90-minute delayed imaging. In our study, 90-minute delayed imaging was superior to immediate imaging in visualizing at least one axillary lymph node (96.6% vs 26.6%, P<0.0001). Additionally, 90-minute delayed imaging was more superior to 15-minute delayed imaging in visualizing at least one axillary lymph node (96.6% vs 80.8%, P<0.0001). Therefore, we believe that 90-minute delayed imaging should be performed as part of the standard imaging protocol in order to consistently and completely image Tc-99m Tilmanocept drainage to sentinel lymph nodes. However, multiple studies are often necessary for both tumor and sentinel lymph node localization on the day of surgery, which can impose a time constraint on the length of each individual study. While 15-minute delayed imaging is not as optimal as 90-minute delayed imaging in reliably visualizing at least one axillary lymph node, it is superior to immediate imaging based on our results and can be performed in lieu of a 90-minute delayed image if time is limited.

There has been controversy over the utility of preoperative lymphoscintigraphy. Studies have shown that using lymphoscintigraphy is accurate, assists with lymph node mapping, and reduces morbidity and costs by avoiding unnecessary axillary lymph node dissection [15,16]. In contrast, a meta-analysis in 2015 reported no significant impact on identification rate using immediate versus delayed imaging [14]. However, the most recent study concluded that preoperative lymphoscintigraphy is associated with a higher sentinel lymph node biopsy identification rate (98.4%) as compared to a lower rate (84%) when lymph nodes are not visualized on lymphoscintigraphy [12]. Therefore, we believe that visualization of lymph nodes is important for surgical planning and 90-minute delayed imaging is the optimal time to visualize at least one sentinel lymph node.

There are several limitations to this study. At our institution, various injection techniques including both intradermal and/or subcutaneous injection of the affected breast was utilized based on surgeon preference. Injection technique may have an influence in localization efficiency impacting our data. Additionally, not every patient included in our study had immediate, 15-minute, and 90-minute imaging again due to our current imaging protocol based on surgeon preference. Therefore, we were not able to compare for each individual, how many lymph nodes were detected at each time point. Given that our study was retrospective, we only have data from immediate, 15-minute, and 90-minute imaging times. Other imaging times (e.g. 60-minutes) can be studied prospectively in the future.

Potential confounding factors might be present that could have impacted the localization efficiency within each individual. Furthermore, a few patients with history of prior surgery including lumpectomy were included in our study. Prior surgery can impact the localization efficiency of the radiotracer due to disrupted lymphatic drainage. These patients all received both immediate and 90 minute imaging, which should have equally impacted both results. However, this could have impacted the comparison between 15-minute delayed and 90-minute delayed imaging. Future areas of investigation include studying whether injection technique affects localization efficiency and determining the likelihood of identifying lymph nodes visualized on lymphoscintigraphy. Additionally, further research can be performed to determine if Tc-99m Tilmanocept can differentiate normal lymph nodes from metastatic lymph nodes. Finally, further studies can determine if the number of lymph nodes detected on lymphoscintigraphy has any effect on clinical sentinel lymph node detection at the time of surgery.

Conclusion

In conclusion, our study shows that lymphoscintigraphy following injection of Tc-99m Tilmanocept should include 90-minute delayed imaging to reliably and adequately map sentinel lymph nodes prior to surgery.

Disclosure of conflict of interest

None.