Background
Molecular breast imaging (MBI) is a nuclear medicine technique that utilizes a specialized gamma camera system optimized for breast imaging. Because MBI relies on the preferential uptake of a radiopharmaceutical in metabolically active cells to distinguish breast abnormalities from normal parenchyma, the sensitivity of this technique is not affected by mammographic breast density and has been shown to improve detection of mammographically occult cancers. In recent studies, supplemental screening with MBI has been shown to detect an additional 7.5 to 16.5 cancers per 1000 women screened [
1].
To date, most studies of MBI have utilized the radiopharmaceutical
99mTc sestamibi, with only limited reports on the use of alternative radiopharmaceuticals for breast imaging [
2‐
6]. Tc-99m sestamibi is primarily used as a perfusion agent for cardiac imaging, and was serendipitously discovered to also accumulate in breast cancer. The biodistribution of Tc-99m sestamibi is such that the relative uptake of Tc-99m sestamibi in breast cells is a factor of 20 lower than in myocardial cells [
7], and hence only a small proportion of administered activity is sequestered in breast tissue. Despite this low uptake, MBI with Tc-99m sestamibi has been shown to be highly sensitive for the detection of breast cancers [
1]. However some tumors are not well-visualized on MBI. In a series of 286 patients with 390 tumor foci, 49 (12%) were occult on MBI; most of the lesions that were occult on MBI with Tc-99m sestamibi were either small (5 mm or less) and potentially below the resolving power of the camera system or located outside the field of view; however, in 17 of 49 (35%), the reason for the absence of Tc-99m sestamibi uptake in the lesion was unknown [
8].
An ideal alternative radiopharmaceutical to Tc-99m sestamibi would have a more favorable biodistribution to breast tissue and offer better visualization of breast tumors through higher uptake in breast malignancies relative to the amount of uptake in normal breast tissue. A radiopharmaceutical of particular interest is the imaging agent
99mTc-NC100692 (also known as Maraciclatide, GE Healthcare), which is a synthetic cyclic peptide with high affinity for specific integrins, such as αvβ3, that are upregulated during angiogenesis [
9]. Angiogenesis is critical for the growth of solid tumors as tumor growth beyond a volume of 1–2 mm
3 requires independent vasculature [
10,
11]. An indirect approach to imaging angiogenesis has focused on radiotracers targeting the integrin αvβ3 receptors which are significantly upregulated in endothelial cells during angiogenesis and are known to be expressed in breast cancer [
12].
The αvβ3 integrin is a membrane-spanning protein that is expressed preferentially on proliferating endothelial cells associated with neovascularization but is absent in quiescent blood vessels [
13,
14]. The binding of
99mTc maraciclatide has been confirmed to be localized to endothelial cells in the regions of angiogenesis [
15], and may provide a promising alternative radiopharmaceutical to
99mTc sestamibi. Bach-Gansmo et al. first reported on the use of a
99mTc-labeled angiogenesis agent (
99mTc NC100692) for the detection of breast cancers using MBI [
16,
17].
The purpose of this study was to compare relative performance of 99mTc sestamibi and 99mTc maraciclatide in patients with suspected breast cancer, using a high-resolution dedicated gamma camera for MBI.
Methods
Study population
This study was performed under an IRB-approved, Health Insurance Portability and Accountability Act-compliant research protocol, and written informed consent was obtained from all participants. As 99mTc maraciclatide is a not an FDA-approved radiopharmaceutical, this study was performed under an investigator IND cross-referred to the IND held by the radiopharmaceutical manufacturer, GE Healthcare Life Sciences (Pittsburgh, PA).
Female patients with known or suspected breast cancer were offered participation in the study. Eligible patients included those with at least one breast lesion identified by clinical findings, mammography, targeted ultrasound, or magnetic resonance imaging that was considered suspicious or highly suggestive of malignancy on the American College of Radiology Breast Imaging Reporting and Data System (BI-RADS) scale [
18] and scheduled for biopsy. MBI examinations were performed prior to breast biopsy when possible (
N = 32). Patients were also eligible to have MBI performed after breast biopsy if the lesion was pathologically proven as malignant and estimated to be at least 1.5 cm in maximum diameter (
N = 7). This requirement on lesion size was to avoid recruitment of patients in whom a significant portion of the lesion would likely be removed at the time of biopsy.
Molecular breast imaging examinations
Patients underwent two MBI examinations—one with
99mTc sestamibi and one with
99mTc maraciclatide. MBI was performed on a dual-head system that comprised two compact cadmium zinc telluride detectors with 1.6 × 1.6 mm pixels (LumaGem system, Gamma Medica, Salem, NH) and was equipped with high sensitivity registered tungsten collimators [
19]. An energy acceptance window of 110–154 keV was used [
20,
21].
When possible, patients were scheduled for 99mTc sestamibi and 99mTc maraciclatide MBI examinations on separate days (2-day protocol; N = 20). If the patient schedule did not permit a 2-day protocol, then a 1-day protocol was followed (N = 19), as described below. For logistical reasons related to preparation and quality control of the 99mTc maraciclatide and to the wait time after injection of the 99mTc maraciclatide, the order of the tests was not randomized. For all patients, MBI with 99mTc sestamibi was performed first, followed by MBI with 99mTc maraciclatide. Between 24–72 h after injection of the 99mTc maraciclatide, patients were contacted to determine if any adverse events (AEs) occurred.
2-day protocol
Patients received an intravenous injection of 300 MBq (8 mCi)
99mTc sestamibi for the first MBI scan on day 1 and 300 MBq (8 mCi)
99mTc maraciclatide for the second MBI scan on day 2, performed at least 16 h after the day 1 examination. Injections were given in the contralateral arm to the breast with the suspected lesion and residual syringe activity was measured after injection. The times at which the injections were performed, imaging was commenced, and dose activities were measured, were recorded to permit accurate correction for decay of the
99mTc and computation of administered activities. Imaging commenced approximately 5 min post-injection with
99mTc sestamibi and approximately 45 min post-injection with
99mTc maraciclatide. The difference in wait time before imaging with
99mTc sestamibi and
99mTc maraciclatide was to allow for known differences in rate of uptake of these radiopharmaceuticals in breast tumors [
16]. Bilateral cranio-caudal (CC) and medio-lateral oblique (MLO) analogous views were obtained under light compression for 10 min per view for a total imaging time of 40 min. The compressed breast thickness for each view acquired during the
99mTc sestamibi MBI examination was recorded and replicated at the
99mTc maraciclatide MBI examination to assist with consistent positioning of the breast between studies.
1-day protocol
The 1-day protocol procedure was similar to that described above, except that after completion of the first MBI scan, performed with 300-MBq (8 mCi) injection of 99mTc sestamibi, patients were immediately injected with 740 MBq (20 mCi) 99mTc maraciclatide for the second MBI scan. Approximately 45 min post-injection of the 99mTc maraciclatide, images were acquired as described above.
The rationale for the higher administered dose of 99mTc maraciclatide for the 1-day protocol was to minimize the contribution of 99mTc sestamibi to the second scan. The time interval from injection of the 99mTc sestamibi to acquisition of the 99mTc maraciclatide images was approximately 90 min for each patient (45 min 99mTc sestamibi injection and imaging, 45 min wait after injection of 99mTc maraciclatide) allowing for physical decay of the 99mTc sestamibi to at least 250 MBq, and physical decay of the 99mTc maraciclatide to 680 MBq. No adjustment was made for any biological washout. Therefore, the relative contribution of counts from the 99mTc maraciclatide and 99mTc sestamibi injections in the second MBI scan was anticipated to be approximately 2.7:1 (~680 MBq 99mTc maraciclatide/~250 MBq 99mTc sestamibi).
Histopathology
Histopathologic classification was by the most severe of findings from surgical excision or by core needle or vacuum-assisted biopsy. Tumor size was obtained from the maximum tumor extent reported on pathology. Molecular subtype of malignant lesions was characterized as luminal (ER+ and HER2−), HER2-enriched (HER2+/ER−/PR−), or triple negative (ER−/PR−/HER2−).
Image analysis
Image manipulation and region of interest (ROI) analysis were performed on the 99mTc maraciclatide and 99mTc sestamibi MBI images using an Xeleris workstation (GE Healthcare, Milwaukee, WI).
Radiopharmaceutical uptake in breast lesions was measured as a lesion-to-background (L/B) ratio, performed as follows. A 3 × 3 median filter was first applied to all images to minimize noise. An ROI was manually drawn to encompass the area of focal uptake corresponding to the lesion using the upper or lower detector CC or MLO view that best visualized the lesion. This lesion ROI was copied to the complementary upper or lower detector image. The maximum count in each lesion ROI was obtained and a geometric mean (GM) of the two counts was calculated. Background activity was measured from a second circumferential ROI that was drawn over a region of normal tissue surrounding the lesion. This background ROI was again copied to the corresponding area of normal tissue on the upper or lower detector image. The average count in each background ROI was obtained and a geometric mean value was calculated. L/B ratio was defined as
$$ \mathrm{L}/\mathrm{B}\ \mathrm{ratio} = \mathrm{G}\mathrm{M}\ \left(\mathrm{maximum}\ \mathrm{counts}\ \mathrm{in}\ \mathrm{lesion}\right)\ /\ \mathrm{G}\mathrm{M}\ \left(\mathrm{average}\ \mathrm{counts}\ \mathrm{in}\ \mathrm{background}\right) $$
For the 1-day protocol studies, a cross-talk correction was applied as follows. Count measurements from the lesion and background ROIs from the
99mTc sestamibi images were decay corrected for the time interval between the
99mTc maraciclatide and
99mTc sestamibi acquisitions and reduced further by a factor of 0.94 (to adjust for biological washout of the
99mTc sestamibi from breast tissue) [
22]. These counts were then subtracted from counts in the
99mTc maraciclatide images. No cross-talk correction was necessary for the 2-day protocol.
MBI interpretation
A breast imaging fellowship-trained radiologist with 4 years of experience interpreting MBI (KNH) performed two independent reading sessions, separated by 4 weeks, of the MBI studies. Each session comprised a random order of
99mTc sestamibi and
99mTc maraciclatide studies; only one study appeared for each patient per session. The radiologist was blinded to the radiopharmaceutical, other imaging findings, and all clinical information including pathology findings. MBI studies were interpreted according to a validated lexicon for gamma imaging of the breast [
23,
24].
The radiologist identified breast lesions and assigned a final assessment on a per-breast basis using a 1 to 5 scale that parallels BI-RADS assessment categories. Assessments were as follows: 1 (negative), 2 (benign), 3 (probably benign), and 4 (suspicious) or 5 (highly suggestive of malignancy) [
23]. Assessments of 3 or higher were considered positive. Lesion size was measured from the CC or MLO views that best demonstrated the full extent of disease. The radiologist also gave an overall assessment of image quality on a 4-category scale (poor, suboptimal, acceptable, good).
Statistical analysis
The proportions of malignant tumors and benign lesions detected by 99mTc maraciclatide and 99mTc sestamibi were compared using McNemar’s test for correlated proportions. A Wilcoxon signed-rank test was used to determine whether assessment scores differed between the 99mTc maraciclatide and 99mTc sestamibi studies.
Tumor extent measured on 99mTc maraciclatide and 99mTc sestamibi imaging was compared using a paired t test. The correlations between tumor extent measured on imaging and lesion extent measured at pathology were determined by Pearson correlation coefficient (r).
L/B ratios of lesions detected on 99mTc maraciclatide and 99mTc sestamibi imaging were compared using paired t-tests. A student’s t-test was used to compare L/B ratio between malignant and benign lesions for each radiopharmaceutical. Correlation in L/B ratio measured on 99mTc maraciclatide and 99mTc sestamibi MBI were also determined by Pearson correlation coefficient (r). Area under the curve (AUC), obtained from ROC analysis of both radiopharmaceuticals, was used to determine the value of the L/B ratio in discriminating between malignant and benign lesions. Correlation of L/B ratios with tumor subtype, considered as three ordinal categories (luminal, HER-2 enriched, triple negative), was determined by Spearman correlation coefficient (r
s
).
Statistical analyses were two sided with a significance level of 0.05.
Discussion
The overall results of this study show that 99mTc maraciclatide offers comparable uptake in breast tumors to 99mTc sestamibi. Lesion extent and L/B ratio measured from the 99mTc maraciclatide images correlated closely with values measured from the 99mTc sestamibi images. Our findings suggest that 99mTc maraciclatide offers only a marginal improvement in sensitivity over 99mTc sestamibi that did not reach significance in this study.
Both radiopharmaceuticals demonstrated comparable uptake in benign lesions. It was hoped that the degree of uptake of the
99mTc maraciclatide in these lesions would be less than that of
99mTc sestamibi, thereby providing better discrimination between benign and malignant lesions. However, factors that contribute to the uptake of
99mTc sestamibi in some benign lesions appear to equally influence the uptake of
99mTc maraciclatide. Previous studies have shown that angiogenesis in breast tissue is initiated at the start of hyperplasia before there is any morphological evidence of malignancy [
25]. Hence, it is not unexpected that there would be uptake of
99mTc maraciclatide in benign lesions. The uptake is similar to that observed with
99mTc sestamibi, thereby offering little or no improvement in discriminating power between benign and malignant lesions. ROC analysis (Fig.
7) showed only a marginal, and non-significant, difference in the ability of the two radiopharmaceuticals to distinguish benign from malignant lesions.
Bach-Gansmo et al. [
15,
16] were the first to report on the high sensitivity of
99mTc maraciclatide for the detection of breast cancer. In their studies, no comparison of the relative uptake of
99mTc maraciclatide with
99mTc sestamibi was performed. Ma et al showed that an analog of maraciclatide,
99mTc RGD, demonstrated marginally better uptake in breast cancers than
99mTc sestamibi, but that difference was not statistically significant. Furthermore they found that
99mTc RGD did not provide any significant advantage over
99mTc sestamibi in distinguishing benign from malignant lesions [
6]. In agreement with the findings of Ma et al, we found a good correlation between the uptakes of the 2 radiopharmaceuticals and no difference in the ability of either radiopharmaceutical to distinguish malignant from benign lesions (Fig.
7). Likewise, apparent lesion extent appeared to be similar with both radiopharmaceuticals.
99mTc sestamibi is currently the only radiopharmaceutical that is FDA-approved for breast imaging [
26]. Several recent studies have shown its clinical use as an imaging agent for the detection of breast cancer in the screening environment [
27‐
29].
99mTc sestamibi also offers a practical advantage over
99mTc maraciclatide in that it does not require a 45-min wait period post injection. With an excellent safety record and no serious adverse events associated with over 30 years of clinical use [
30],
99mTc sestamibi has essentially become the primary radiotracer used in the majority of MBI studies. Uptake of
99mTc sestamibi in tumors is primarily dependent on blood flow to the tumor bed and on increased uptake in the mitochondria of the tumor cells [
31]. As such, it has served as a general-purpose imaging agent for the detection of breast cancer, but these same characteristics also result in uptake in benign lesions and conditions such as fibroadenomata, papillomas, areas of inflammation, necrosis, and benign fibrocystic changes [
32,
33].
99mTc maraciclatide may offer some advantages over
99mTc sestamibi for breast imaging.
99mTc sestamibi is known to have significantly lower uptake in invasive lobular carcinomas than in invasive ductal carcinomas [
8]. Figure
1 would indicate that
99mTc maraciclatide may be a better imaging agent than
99mTc sestamibi for the detection of invasive lobular carcinomas. However, the same figure shows that
99mTc maraciclatide can occasionally overestimate disease. The reason for this overestimation is unknown but may reflect hyperplasia in the breast tissue surrounding the lesion.
99mTc maraciclatide is extracted primarily through the hepatobiliary system with no uptake in the myocardium. In some patients, this may result in better breast image quality as the absence of myocardial activity adjacent to the breast may result in less scatter in the breast images. This point is well demonstrated in Fig.
5 where increased activity close to the chest wall is evident on the
99mTc sestamibi image but absent from the
99mTc maraciclatide image. Both radiopharmaceuticals have a comparable radiation dosimetry profile. The effective radiation dose from
99mTc maraciclatide is 7.8 uSv/MBq [
34] which is comparable to that from
99mTc sestamibi (7.1 uSv/MBq) [
35]. However, no efforts have been expended to determine if the uptake of
99mTc maraciclatide in breast tissue could be modulated through patient preparation as has been done with
99mTc sestamibi [
22].
One potential role for breast imaging with MBI is in patients undergoing neoadjuvant chemotherapy (NAC) where early prediction of response to NAC offers a potential opportunity to change treatment strategy in cases of inadequate response. Previous studies have shown that quantitative assessment of tumor uptake of
99mTc sestamibi demonstrated the ability to differentiate between pathological responders and non-responders as early as 3–5 weeks after initiation of NAC [
36]. To date, there is little understanding of what these changes in tumor uptake represent. Our findings did not support an association between the uptake of either radiopharmaceutical and molecular subtype. While the Luminal cancers had some of the lowest L/B ratios, there was no clear relationship between uptake and molecular subtype. Additional studies will be needed to determine what relationship, if any, exists between the degree of uptake of either radiopharmaceutical in a tumor and molecular subtype or tumor histopathology.
This study had some limitations. The sample size of 40 patients limits our ability to detect small differences in the sensitivity of the two radiopharmaceuticals. A second limitation was the use of a 1-day protocol in approximately half of patients, imposed by limitations in patient’s availability and inability to return for a 2-day protocol. Figure
6b showed no evidence that L/B ratios measured using the 1-day protocol were biased high or low relative to L/B ratios measured using the 2-day protocol. However, in those patients who underwent the 1-day protocol, image quality with the
99mTc maraciclatide images may have been better than that observed in the
99mTc sestamibi images due to the increased administered dose and corresponding reduced noise in the clinical images. All of the MBI examinations acquired for this study were considered to be of adequate image quality for diagnostic interpretation and appropriate adjustments to image counts were made for ROI analyses. The order of MBI examinations with the two radiopharmaceuticals was not randomized due to restrictions with preparation of the
99mTc maraciclatide; however, MBI studies were reviewed in a randomized order with the radiologist blinded to radiopharmaceutical and all clinical information in order to avoid biasing the interpretation.