Background: Background parenchymal uptake (BPU), which refers to the level of Tc-99m sestamibi uptake within normal fibroglandular tissue on molecular breast imaging (MBI), has been identified as a breast cancer risk factor, independent of mammographic density. Prior analyses have used subjective categories to describe BPU. We evaluate a new quantitative method for assessing BPU by testing its reproducibility, comparing quantitative results with previously established subjective BPU categories, and determining the association of quantitative BPU with breast cancer risk. Methods: Two nonradiologist operators independently performed region-of-interest analysis on MBI images viewed in conjunction with corresponding digital mammograms. Quantitative BPU was defined as a unitless ratio of the average pixel intensity (counts/pixel) within the fibroglandular tissue versus the average pixel intensity in fat. Operator agreement and the correlation of quantitative BPU measures with subjective BPU categories assessed by expert radiologists were determined. Percent density on mammograms was estimated using Cumulus. The association of quantitative BPU with breast cancer (per one unit BPU) was examined within an established case-control study of 62 incident breast cancer cases and 177 matched controls. Results: Quantitative BPU ranged from 0.4 to 3.2 across all subjects and was on average higher in cases compared to controls (1.4 versus 1.2, p < 0.007 for both operators). Quantitative BPU was strongly correlated with subjective BPU categories (Spearman’s r =0.59 to 0.69, p < 0.0001, for each paired combination of two operators and two radiologists). Interoperator and intraoperator agreement in the quantitative BPU measure, assessed by intraclass correlation, was 0.92 and 0.98, respectively. Quantitative BPU measures showed either no correlation or weak negative correlation with mammographic percent density. In a model adjusted for body mass index and percent density, higher quantitative BPU was associated with increased risk of breast cancer for both operators (OR = 4.0, 95% confidence interval (CI) 1.6–10.1, and 2.4, 95% CI 1.2–4.7). Conclusion: Quantitative measurement of BPU, defined as the ratio of average counts in fibroglandular tissue relative to that in fat, can be reliably performed by nonradiologist operators with a simple region-of-interest analysis tool. Similar to results obtained with subjective BPU categories, quantitative BPU is a functional imaging biomarker of breast cancer risk, independent of mammographic density and hormonal factors. Keywords: Breast density, Breast cancer risk, Molecular breast imaging, Tc-99m sestamibi, Mammography * Correspondence: email@example.com Department of Physiology and Biomedical Engineering, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Hruska et al. Breast Cancer Research (2018) 20:46 Page 2 of 10 Background a lackof uptakeinfibroglandulartissue (photopenic) to Mammographic density, or the amount of fibroglandular very high intensity uptake (marked), as shown in Fig. 1. tissue in the breast as depicted on a mammogram, is Importantly, subjective categories of high BPU were known to reduce the accuracy of mammography in detect- found to be associated with risk of incident breast ing cancer [1–3]. Density is also independently associated cancer relative to those with low BPU in a case-control with breast cancer risk as established by numerous analysis (odds ratio (OR) range from 3 to 5) after adjust- analyses conducted over the last 40 years, consistently ment for mammographic density and exogenous hormone showing women with the densest breasts to be four- to use . These results suggest that BPU is a functional six-times more likely to be diagnosed with breast cancer imaging biomarker that depicts risk-related aspects of compared with those with low density [4, 5]. However, be- fibroglandular tissue not observed through measures of cause breast density is highly prevalent (approximately 40 mammographic density alone. to 50% of screening-eligible women have heterogeneously Prior investigations of BPU on MBI have used a or extremely dense breasts according to American College subjective measure which includes the following four of Radiology (ACR) Breast Imaging-Reporting and Data categories: photopenic, minimal to mild, moderate, and System (BI-RADS) categories [6, 7]), it is impractical for marked [11, 12]. While expert readers were observed to clinicians to consider all women with dense breasts to be have substantial agreement using this subjective classifi- at elevated risk since doing so would warrant consider- cation (κ = 0.84) , only fair agreement was observed ation of supplemental screening or preventive options in among nonexpert readers (κ = 0.31) . A quantitative nearly half the screening population. To identify the sub- tool for BPU assessment may improve reproducibility of set of women with dense breasts at greatest risk of breast the measure. Additionally, a quantitative measure of cancer, and those most likely to benefit from these strat- BPU on a continuous scale would provide a more pre- egies, improved risk stratification tools are needed. cise measurement and could therefore more accurately Molecular breast imaging (MBI) is a nuclear medicine monitor changes in BPU over serial MBI examinations. test that uses dedicated gamma cameras and injection of The objective of this work was to develop and evaluate a radiotracer, typically Tc-99m sestamibi, to detect breast a new quantitative method for assessing BPU by asses- cancer. As MBI is a functional imaging technique that sing its reproducibility, comparing quantitative to cat- relies on the preferential uptake of radiotracer in meta- egorical BPU measures, and determining the association bolically active cells, it is able to reveal cancers obscured of quantitative BPU with breast cancer risk. by breast density on mammography. MBI can also depict the functional uptake of radiotracer in benign fibro- Methods glandular tissue, which has been termed background Study population parenchymal uptake (BPU). High levels of BPU are hy- This retrospective analysis was compliant with the US pothesized to represent breast tissue with elevated meta- Health Insurance Portability and Accountability Act and bolic activity due to a combination of factors such as approved by the Mayo Clinic Institutional Review Board, abundant mitochondria, cellular proliferation, and blood which issued a waiver of informed consent. A case-control flow [8, 9]. Among women with similar mammographic study previously established to evaluate the association of density, BPU has been observed to vary substantially from subjective BPU categories with breast cancer risk was used Fig. 1 BPU subjective categories. Example MBI images from four women, all acquired in right mediolateral oblique projection, showing the range of BPU observed: a photopenic BPU, b minimal to mild BPU, c moderate BPU, and d marked BPU Hruska et al. Breast Cancer Research (2018) 20:46 Page 3 of 10 in the current analysis . As previously described, we Images established this case-control group within a cohort of MBI examinations were performed as previously described patients who underwent MBI between 1 February 2005 . Briefly, MBI was performed using a dedicated dual- and 28 February 2014 (n = 3202) and who were followed head gamma camera system equipped with semiconductor- for breast cancer diagnoses through 31 December 2014. A based detectors (cadmium zinc telluride). Following total of 3027 patients were eligible for study inclusion as injection of Tc-99m sestamibi, two-view acquisitions they had provided general authorization to use medical re- (craniocaudal (CC) and mediolateral oblique (MLO)) of cords in research, did not have breast implants at the time each breast were made. Thus, the entire MBI dataset com- of MBI, and did not have a prior history of breast cancer prised of eight images: a CC and MLO projection of left or were diagnosed at the time of MBI (within 180 days). and right breasts acquired with two detector heads of the Any subject with a diagnosis of invasive breast cancer or dual head system (Fig. 2). ductal carcinoma in situ (DCIS) 180 days or more following Mammograms performed closest to the time of MBI the MBI was considered a case. Sixty-two cases were identi- were used for density analysis. The median time from fied; the median time from MBI to diagnosis was 3.1 years. mammogram to MBI was 0 days (interquartile range (IQR) Control subjects were matched to cases on age (within 0–1 days). Mammographic density was classified according 5 years), menopausal status, year of MBI, and follow-up to the ACR BI-RADS breast composition categories (4th interval (at least as long as matched case); the median edition) at the time of clinical interpretation . Density follow-up time was 6.1 years. Up to three controls per case was also quantitatively measured as percent density (PD) by were originally selected (n =179). Two controls were ex- a trained operator using a semi-automated software tool cluded as their mammograms were unavailable for use in (Cumulus; University of Toronto, Toronto, ON, Canada the quantitative BPU software program (described below) ), as previously described . leaving 177 controls for the current analysis. Patient information at the time of MBI, including body Quantitative BPU measurement mass index (BMI), menopausal status, postmenopausal The subjective BPU categories, as defined in a validated hormone use, breast biopsy history, and family history of lexicon for gamma imaging of the breast [11, 12], are breast cancer, was obtained from research study ques- intended to describe the relative intensity of radiotracer up- tionnaires and medical record review. take in normal breast parenchyma (or fibroglandular tissue) Fig. 2 Example layout of images for quantitative BPU region-of-interest analysis. Bilateral mammogram and MBI views in CC and MLO projections are displayed. Mammogram in top row (from left to right) comprises right CC, left CC, right MLO, and left MLO projections. The same projections acquired by MBI are shown for the upper detector (middle row) and for the lower detector (bottom row) Hruska et al. Breast Cancer Research (2018) 20:46 Page 4 of 10 compared with intensity of uptake in subcutaneous fat. area predominantly made of fat and one to include pre- These categories and their definitions are as follows: 1) dominantly fibroglandular tissue. The operator then manu- photopenic BPU, fibroglandular intensity less than fat inten- ally adjusted the fibroglandular ROI using an intensity sity; 2) minimal to mild BPU, fibroglandular intensity equal threshold to reject less-dense tissue, thereby reducing the to or slightly greater than fat intensity; 3) moderate BPU, overall size of the ROI and making it more specific to fibroglandular intensity greater than mild but less than twice dense tissue. Finally, the three mammogram ROIs were as intense as fat; and 4) marked BPU, fibroglandular inten- copied to the corresponding MBI views and manually sity greater than twice fat intensity. The quantitative BPU scaled and rotated as needed (all three as a single object) tool was designed to provide values that reflect these defini- to register the outer outline of the mammogram to the tions, such that it measures the ratio of average image MBI breast outline. This process was done to account for counts (counts per pixel) in fibroglandular tissue versus fat. differences in breast position and compression force be- As MBI examinations create functional images of tween the mammogram and MBI images. Example ROIs radiotracer localization, they do not provide distinct ana- are shown in Fig. 3. The ratio of average counts (counts tomic landmarks of the breast from which to distinguish per pixel) in the fibroglandular ROI versus fat ROI was fibroglandular tissue and fat. However, as MBI is taken as a measure of quantitative BPU. acquired in positions analogous to mammography (CC Two operators used this program to perform quantita- and MLO), a mammogram viewed in conjunction with tive BPU measurements on each of the eight MBI views MBI may be used to generally determine fibroglandular for all subjects. Each operator independently performed and fat locations. In this first approach to develop a quan- each step in the ROI process as described above, while titative tool for BPU, we used a corresponding mammo- blinded to the other operator’s results and blinded to gram to identify fibroglandular and fat regions, applied patient identity and case status. One operator was a these regions to the MBI images, and determined the novice to medical imaging and image processing tools fibroglandular-to-fat count ratio on MBI as a quantitative and the other operator was a nuclear medicine technolo- BPU value, described in more detail as follows. gist familiar with mammography, MBI, and image pro- An in-house software program was created to allow cessing software. To determine intraoperator agreement, simultaneous display of MBI examinations with mam- one operator performed quantitative BPU measurements mograms, as shown in Fig. 2. Using this program, each a second time on a sample of 48 subjects. mammogram and its corresponding MBI images were processed as follows. First, a region of interest (ROI) out- Statistical analysis lining the entire breast visible on the mammogram was Case and control characteristics are summarized using automatically drawn, based on a manually adjustable inten- frequencies for categorical variables or mean and stand- sity threshold. Next, two other ROIs were drawn by the ard deviation (SD) with range for continuous variables. operator on the mammogram view—one to encompass an Conditional logistic regression was used to evaluate the Fig. 3 Example ROIs for quantitative BPU assessment. Breast images acquired in the mediolateral oblique position for two patients. Fibroglandular tissue is defined by the orange ROI and fat is defined by the green ROI. In panels a and b,the mammogram (a) and MBI (b) are shown for a patient classified as having photopenic BPU, measured quantitatively as 0.4. In panels c and d, the mammogram (c) and MBI (d) are shown for a patient with marked BPU, measured quantitatively as 2.4 Hruska et al. Breast Cancer Research (2018) 20:46 Page 5 of 10 primary hypothesis that quantitative BPU is associated 0.90–0.94) and ICCs ranged from 0.75 (95% CI 0.68–0.80) with breast cancer. Models were adjusted for BMI and to 0.92 (95% CI 0.86–0.95) across the eight views. PD and postmenopausal hormone use. All ORs reported Intraoperator agreement in quantitative BPU was 0.98 (0.96– are for a one-unit change in quantitative BPU measure- 0.99) for the average of the eight views and ICCs ranged ment. Analyses were repeated within premenopausal and from 0.80 (95% CI 0.67–0.88) to 0.91 (95% CI 0.84–0.95) postmenopausal subgroups. across the eight views. Agreement in quantitative BPU across eight MBI views Quantitative BPU measurements correlated well with was assessed using several methods including intraclass subjective BPU categories previously assessed by radiolo- correlation (ICC), principal component (PC) analysis, gists , as shown in Fig. 4. Spearman correlations and a nested random effects model to inform the sum- ranged from 0.59 to 0.69 (all p values < 0.0001) for the mary quantitative measure. ICC for each of the eight four combinations of two categorical BPU readers and views was calculated and ranged from 0.74 to 0.92. PC two quantitative BPU operators. analysis revealed a lack of multidimensionality among Quantitative BPU measures and PD showed weak or the eight views, as the first and second PCs explained 0.83 no correlation (all p values > 0.05 unless specified) in the and 0.09 of the variation in all eight measures. Lastly, a total sample (operator 1: r = −0.11, operator 2: r = −0.07) nested random effects model demonstrated that only 2% (Fig. 5), or separately in cases (operator 1: r = 0.12, operator of the variation in quantitative BPU was due to the eight 2: r = 0.14), or controls (operator 1: r = −0.20, p = 0.009; views within a subject and 83% was intersubject variation, operator 2: r = −0.13). leaving 15% of variation due to random error. Therefore, with low dimensionality and limited variation across the Breast cancer cases versus controls eight images, we used an average of the quantitative BPU Quantitative BPU was associated with breast cancer risk in values from the eight views as the quantitative BPU meas- models adjusted for BMI, with odds ratios per 1 unit BPU ure for each subject herein. Analyses of individual views of 3.70 (95% CI 1.54–8.92) and 2.37 (95% CI 1.19–4.70) and other combinations (averages by breast side and view) for the two operators, respectively (Table 3). Results were along with their p values and area under the curves similar for models including PD or postmenopausal hor- (AUCs) are summarized by operator in Additional file 1 mone use. (Table S1). In analyses limited to postmenopausal women, quanti- Interoperator and intraoperator agreement in quantita- tative BPU remained associated with breast cancer risk tive BPU measures were summarized by ICC. Agreement (BMI-adjusted OR per 1 unit BPU = 5.57, 95% CI 1.62–19.08, between quantitative BPU and subjective BPU categories for operator 1, and 2.91, 95% CI 1.15–7.35, for oper- and PD measured on mammography was determined by ator 2); however, the BPU measure was not a statisti- Spearman and Pearson correlations, respectively. For cally significant predictor for breast cancer in the all comparisons, p < 0.05 was considered statistically premenopausal subset for either operator in a small significant. Analyses were performed within SAS (Cary, sample of 13 cases. NC; version 9.4). Analysis of the eight MBI views separately by operator, as well as averages of right breast, left breast, MLO views, Results CC views, upper detector views, and lower detector views Subject characteristics were considered (Additional file 1: Table S1). All models A total of 239 subjects, including 62 cases and 177 con- concluded that higher quantitative BPU was significantly trols, were included in the study, with their characteristics associated with breast cancer risk with the exception of presented in Table 1. Cases and controls were similar on one view under operator 2 (OR = 1.5, p = 0.18). ORs matched variables of age and menopausal status, as ex- ranged 2.0 to 4.5 for operator 1 and from 1.5 to 2.6 for pected. Cases and controls were not significantly different operator 2, but reliably showed consistent overall model in any other characteristic examined, including BI-RADs performance with AUCs from 0.57 to 0.62. density and PD. Quantitative BPU measures ranged from 0.41 to 3.18, with mean (IQR) values of 1.36 (1.09–1.54) Discussion for cases versus 1.18 (0.97–1.31) for controls for operator In this first evaluation of a simple region-of-interest tool 1(p = 0.002) and 1.36 (1.04–1.56) for cases and 1.19 (0.93– for obtaining quantitative measurements of BPU on 1.33) for controls for operator 2 (p =0.007) (Table 2). MBI, we found an association of quantitative BPU measurements with breast cancer risk, similar to that BPU agreement observed in a prior analysis of subjective BPU categories. The two operators showed good agreement in assessing This association was independent of mammographic quantitative BPU; the interoperator ICC for the average density. In fact, in line with our previous observation of the eight views was 0.92 (95% confidence interval (CI) that BPU can vary widely among women with similar Hruska et al. Breast Cancer Research (2018) 20:46 Page 6 of 10 Table 1 Characteristics of breast cancer cases and controls, matched on age and menopausal status Characteristic Breast cancer cases (n = 62) Controls (n = 177) P Age at MBI (years) 60.3 ± 10.6 (38–88) 60.2 ± 10.4 (38–86) NA Menopausal status NA Premenopausal 13 (21) 38 (22) Postmenopausal 49 (79) 138 (78) 2 a Body mass index (kg/m ) 27.7 ± 6.4 (18.8–55.5) 26.2 ± 4.7 (18.6–44.3) 0.06 Postmenopausal systemic hormone therapy 0.57 Current use at MBI 13 (27) 44 (31) No current use at MBI 36 (73) 97 (69) BI-RADS density 0.81 Almost entirely fat 1 (2) 3 (2) Scattered fibroglandular densities 10 (16) 34 (19) Heterogeneously dense 44 (71) 113 (64) Extremely dense 7 (11) 26 (15) Percent density 24.8 ± 8.3 (3.5–48.0) 24.6 ± 10.2 (1.8–53.8) 0.93 Tumor invasiveness Invasive 45 (73) NA DCIS 17 (27) NA Gail model 5-year risk 2.7 ± 1.5 (0.6–7.2) 2.4 ± 1.5 (0.5–9.5) 0.23 BCSC model 5-year risk 2.6 ± 1.2 (0.7–5.4) 2.3 ± 1.5 (0.4–13.2) 0.29 Family history of breast cancer 0.45 One or more first-degree relatives 33 (53) 86 (48) No first-degree relatives 29 (47) 93 (52) Personal history of biopsy showing atypia or LCIS 0.07 Yes 6 (10) 6 (3) No 56 (90) 173 (97) Unless otherwise noted, data are number of patients and data in parentheses are percentages Data are mean ± standard deviation; data in parentheses are the range Data are among postmenopausal women only (49 breast cancer cases; 138 controls) BCSC, Breast Cancer Surveillance Consortium; BI-RADS, Breast Imaging-Reporting and Data System; DCIS, ductal carcinoma in situ; LCIS, lobular carcinoma in situ; MBI, molecular breast imaging; NA, not available mammographic density [10, 13], we saw no association measures describe the amount of fibroglandular tissue in between quantitative BPU and quantitative percent dens- the breast by its anatomic appearance, BPU describes the ity in the current analysis. functional radiotracer uptake within that fibroglandular The lack of relationship between BPU and mammo- tissue relative to the uptake in fat. Furthermore, density graphic density is not unexpected since BPU and density assessment tools, such as Cumulus, use a binary decision are fundamentally different imaging features. While density to categorize image pixels of a mammogram as “dense” or Table 2 Molecular breast imaging (MBI) quantitative background parencymal uptake (BPU) components by case status Characteristic Breast cancer cases Controls Operator n =62 n = 176 P Average counts in fibroglandular tissue, mean ± SD (range) 1 41.5 ± 20.7 (6.8–116.6) 37.6 ± 23.2 (7.3–218.4) 0.08 2 41.2 ± 21.1 (7.2–194.9) 36.5 ± 21.7 (6.5–114.6) 0.08 Average counts in fat, mean ± SD (range) 1 31.5 ± 14.3 (5.9–87.4) 32.7 ± 17.2 (6.2–95.7) 0.99 2 32.6 ± 16.0 (5.7–90.7) 32.5 ± 17.0 (7.2–98.9) 0.78 Quantitative BPU (fibroglandular/fat), mean ± SD (range) 1 1.4 ± 0.4 (0.8, 2.8) 1.2 ± 0.3 (0.4, 2.9) 0.002 2 1.4 ± 0.5 (0.8, 2.9) 1.2 ± 0.4 (0.5, 3.2) 0.007 Hruska et al. Breast Cancer Research (2018) 20:46 Page 7 of 10 Fig. 4 Quantitative background parencymal uptake (BPU) measurements by subjective BPU category. BPU as assessed by a) operator 1 versus radiologist reader 1, b) operator 2 versus radiologist reader 1, c) operator 1 versus radiologist reader 2, and d) operator 2 versus radiologist reader 2 with corresponding Spearman correlations (all p values< 0.0001) “non-dense”, and output the proportion of dense pixels, very dense and have low BPU. The quantitative BPU value but do not take into account the intensity of those dense can vary substantially, even when percent density is similar. pixels. In contrast, BPU as measured on MBI is determined For instance, as seen in Fig. 5, women with percent density by the average intensity of the pixels in fibroglandular tis- of about 40% were found to vary in quantitative BPU sue relative to the intensity of pixels in fat. Therefore, it is values from 0.4 to 3.2. possible for a breast to have a small amount of dense tissue The underlying etiology relating BPU of Tc-99m sesta- and yet have high uptake within that dense tissue on MBI, mibi and risk of breast cancer is not yet known. In fact, resulting in high BPU. It is also possible for a breast to be the mechanism of Tc-99m sestamibi uptake in the breast Fig. 5 Quantitative background parencymal uptake (BPU) measurements by mammographic percent density (PD). BPU for breast cancer (BC) cases and controls are shown for both operators, with corresponding Spearman correlations Hruska et al. Breast Cancer Research (2018) 20:46 Page 8 of 10 Table 3 Association of quantitative background parencymal uptake (BPU) (per 1 unit BPU) with breast cancer OR (95% CI), adjusted for BMI OR (95% CI), adjusted for BMI and PD OR (95% CI), adjusted for BMI and postmenopausal hormones Overall Operator 1 3.70 (1.54–8.92) 3.98 (1.58–10.05) 3.85 (1.58–9.38) p value 0.0036 0.0034 0.0030 AUC (95% CI) 0.63 (0.56–0.71) 0.66 (0.59–0.73) 0.63 (0.56–0.70) Operator 2 2.37 (1.19–4.70) 2.35 (1.17–4.71) 2.31 (1.17–4.55) p value 0.0136 0.0163 0.0154 AUC (95% CI) 0.58 (0.51–0.66) 0.61 (0.54–0.69) 0.61 (0.54–0.69) Postmenopausal women (n = 187: 49 cases, 138 controls) Operator 1 5.57 (1.62–19.08) 8.39 (2.10–33.55) 5.87 (1.69–20.36) p value 0.0063 0.0026 0.0053 AUC (95% CI) 0.65 (0.57–0.73) 0.70 (0.62–0.78) 0.65 (0.57–0.73) Operator 2 2.91 (1.15–7.35) 3.62 (1.34–9.79) 2.99 (1.18–7.57) p value 0.0239 0.0113 0.0212 AUC (95% CI) 0.57 (0.48–0.65) 0.66 (0.58–0.74) 0.61 (0.53–0.69) Premenopausal women (n = 51: 13 cases, 38 controls) Operator 1 2.42 (0.73–8.04) 2.09 (0.64–6.79) NA p value 0.1492 0.2224 AUC (95% CI) 0.57 (0.41–0.73) 0.58 (0.42–0.73) Operator 2 1.70 (0.63–4.58) 1.47 (0.54–3.98) NA p value 0.2922 0.4468 AUC (95% CI) 0.62 (0.47–0.78) 0.61 (0.45–0.76) AUC, area under the curve; BMI, body mass index; CI, confidence interval; NA, not applicable; OR, odds ratio; PD, percent density by Cumulus software in general is not well understood. Tc-99m sestamibi was current study, quantitative BPU was strongly associated developed as a tracer for imaging myocardial perfusion with breast cancer in postmenopausal women, but this and was only incidentally discovered to accumulate in association was somewhat attenuated with adjustment breast lesions in women undergoing cardiac testing . for hormone therapy use. In premenopausal women, Tc-99m sestamibi is known to be mostly sequestered in BPU can fluctuate with the menstrual cycle, with higher cellular mitochondria . In breast cancer, its uptake is levels of BPU observed in the luteal phase compared with thought to reflect both blood flow to the tumor and the follicular phase . When we restricted analysis to mitochondrial status, which is affected by the cellular postmenopausal women in the current work, the associ- proliferation rate and apoptotic index [8, 9]. Benign ation of quantitative BPU with breast cancer remained, breast lesions that are highly proliferative, such as atyp- suggesting that the association is not merely reflecting ical lesions and fibroadenomas, can also demonstrate changes in BPU with the menstrual cycle. We did not ob- high uptake of Tc-99m sestamibi that mimics breast serve a significant association in premenopausal women; cancer . Although the etiology of variations in BPU however, the analysis was limited in power due to smaller among women has not been established, it can be hy- numbers (n = 13 cases). pothesized that breast fibroglandular tissue with higher This study found that quantitative BPU assessed by blood flow and more proliferative cells would also ex- operators correlates well with subjective BPU categories hibit higher BPU, and thus may represent tissue that is assessed by expert radiologists. We also found good primed for breast cancer development. agreement in quantitative BPU measurements between Hormonal factors which are known to impact tissue the two operators, one of whom was a novice to medical perfusion and proliferation have been found to impact imaging, indicating that the quantitative method is BPU. We have previously shown that high (moderate or robust and generalizable to other operators. Importantly, marked) BPU is more prevalent among premenopausal our results on the evaluation of eight views showed that women compared with postmenopausal women . In the quantitative BPU obtained from any of the eight MBI postmenopausal women, those using exogenous hormo- views or any of the reported averages of multiple views is nal therapy are more likely to have high BPU . In the a reliable predictor of increased breast cancer risk, shown Hruska et al. Breast Cancer Research (2018) 20:46 Page 9 of 10 under two different operators with varying experience. Conclusions Thus, future investigations could use a single view or Quantitative measurement of BPU, which can be reliably combination of views for quantitative BPU assessment. assessed by nonradiologist operators with a simple MBI is indicated for women with dense breasts, as region-of-interest analysis tool, correlates well with sub- reflected in our study population here where a majority jective BPU categories assessed by expert radiologists. of cases (82%) and controls (79%) were considered mam- Similar to findings with subjective BPU categories, mographically dense. In our institution’s practice, MBI is quantitative BPU measurement is associated with primarily used as a screening tool and is offered to breast cancer risk, independent of mammographic women with dense breasts who seek supplemental density and hormonal factors. These results suggest screening but either do not wish to undergo or do not that quantitative measures of BPU could serve as an meet the high-risk criteria (20% lifetime risk by familial additional tool for identifying a subset of women with models) for screening breast magnetic resonance (MR) mammographically dense breasts who are at greatest imaging. Supplemental screening MBI, performed with risk of breast cancer. reduced administered doses of 300 MBq (8 mCi) Tc- 99m sestamibi, offers a reported incremental cancer detec- Additional file tion rate of 7.7 to 8.8 cancers per 1000 women screened [20, 21]. Although breast density is well-established as a Additional file 1: Table S1. Association of quantitative BPU with breast cancer for each MBI view and combinations of views. (DOCX 18 kb) breast cancer risk factor, for women with dense breasts in- cluded in this study there was no association between breast cancer and mammographic density assessed by Abbreviations ACR: American College of Radiology; AUC: Area under the curve; mammographic categories or quantitative percent density. BI-RADS: Breast Imaging-Reporting and Data System; BMI: Body mass index; Thus, BPU will be an important risk factor for the dense BPU: Background parenchymal uptake; CC: Craniocaudal; CI: Confidence breast population and may offer additional image-based interval; DCIS: Ductal carcinoma in situ; ICC: Intraclass correlation; IQR: Interquartile range; MBI: Molecular breast imaging; MLO: Mediolateral risk information beyond density alone. Further work is oblique; OR: Odds ratio; PC: Principal component; PD: Percent density; needed to determine the impact of incorporating BPU into ROI: Region of interest existing risk models. Although our quantitative BPU method is relatively sim- Funding ple and easy to implement with minimal operator training, This work was supported by grants from the National Cancer Institute (R21 CA 197752), National Center for Advancing Translational Sciences the method does have some limitations. First, given this was (UL1 TR000135), the Mayo Clinic Cancer Center, Fraternal Order of Eagles the first study relating the quantitative BPU to breast cancer, Cancer Research Fund, and the Mayo Clinic Department of Radiology. our estimates for the strength of the association were impre- cise. This can be evidenced by the differential estimates Availability of data and materials The datasets analyzed during the current study are available from the of risk between operators (e.g., OR = 5.87 versus 2.99). corresponding author on reasonable request. These confidence intervals for the estimates are wide and overlapping. Further work is needed to develop a Authors’ contributions comprehensive model on a larger set of patients to CBH, CGS, and CMV conceived and designed the study. JRG, CGS, and REC performed statistical analysis and interpretation of the data. ALC and DHW ensure the risk estimates are properly calibrated. performed the reading of molecular breast imaging examinations and Second, our method for measuring quantitative BPU mammograms. CBH, TNS, ANM, DSL, and AM developed and tested the currently requires user interaction to manually segment quantitative BPU tool. DJR and MKO provided image and questionnaire data from molecular breast imaging research trials. All authors contributed and align regions from the mammogram to the MBI. to drafting and revising the manuscript, and all read and approved the Best results are expected when the breast is similarly final manuscript. positioned on the mammogram and MBI, which is not always possible as they are acquired under separate ex- Ethics approval and consent to participate This retrospective analysis was compliant with the US Health Insurance aminations with the MBI performed under substantially Portability and Accountability Act and approved by the Mayo Clinic less breast compression. Also, our current quantitative Institutional Review Board, which issued a waiver of informed consent. measure is based on the ratio of average pixel intensities in fibroglandular and fat regions obtained in two- Competing interests CBH and MKO receive royalties for licensed technologies per agreement dimensional planar images. Similar to findings from between Mayo Clinic and Gamma Medica, a manufacturer of molecular studies of mammographic density and breast cancer risk, breast imaging systems. The remaining authors declare that they have no a more precise or more reproducible risk association competing interests. may be obtained if the BPU area is considered or volu- metric BPU estimates are made. Future iterations of this Publisher’sNote method are anticipated to be automated and to evaluate Springer Nature remains neutral with regard to jurisdictional claims in published additional factors such as BPU volume. maps and institutional affiliations. Hruska et al. Breast Cancer Research (2018) 20:46 Page 10 of 10 Author details 20. Rhodes DJ, Hruska CB, Conners AL, Tortorelli CL, Maxwell RW, Jones KN, Department of Physiology and Biomedical Engineering, Mayo Clinic, 200 Toledano AY, O'Connor MK. Journal club: molecular breast imaging at First Street SW, Rochester, MN 55905, USA. Department of Health Sciences reduced radiation dose for supplemental screening in mammographically Research, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA. dense breasts. Am J Roentgenol. 2015;204(2):241–51. Department Medicine, Mayo Clinic, 200 First Street SW, Rochester, MN 21. Shermis RB, Wilson KD, Doyle MT, Martin TS, Merryman D, Kudrolli H, 55905, USA. Brenner RJ. Supplemental breast cancer screening with molecular breast imaging for women with dense breast tissue. Am J Roentgenol. 2016;207:1–8. Received: 18 December 2017 Accepted: 27 April 2018 References 1. Carney PA, Miglioretti DL, Yankaskas BC, Kerlikowske K, Rosenberg R, Rutter CM, Geller BM, Abraham LA, Taplin SH, Dignan M, et al. Individual and combined effects of age, breast density, and hormone replacement therapy use on the accuracy of screening mammography. Ann Intern Med. 2003; 138(3):168–75. 2. Mandelson MT, Oestreicher N, Porter PL, White D, Finder CA, Taplin SH, White E. Breast density as a predictor of mammographic detection: comparison of interval- and screen-detected cancers. J Natl Cancer Inst. 2000;92(13):1081–7. 3. Buist DS, Porter PL, Lehman C, Taplin SH, White E. Factors contributing to mammography failure in women aged 40-49 years. J Natl Cancer Inst. 2004; 96(19):1432–40. 4. Vachon CM, Gils CH, Sellers TA, Ghosh K, Pruthi S, Brandt KR. Mammographic density, breast cancer risk and risk prediction. Breast Cancer Res. 2007;9(6):217. 5. McCormack VA. Breast density and parenchymal patterns as markers of breast cancer risk: a meta-analysis. Cancer Epidemiol Biomark Prev. 2006; 15(6):1159–69. 6. Brandt KR, Scott CG, Ma L, Mahmoudzadeh AP, Jensen MR, Whaley DH, Wu FF, Malkov S, Hruska CB, Norman AD, et al. Comparison of clinical and automated breast density measurements: implications for risk prediction and supplemental screening. Radiology. 2016;279(3):710–9. 7. Sprague BL, Gangnon RE, Burt V, Trentham-Dietz A, Hampton JM, Wellman RD, Kerlikowske K, Miglioretti DL. Prevalence of mammographically dense breasts in the United States. J Natl Cancer Inst. 2014;106(10) 8. Arbab AS, Koizumi K, Toyama K, Araki T. Uptake of technetium-99m- tetrofosmin, technetium-99m-MIBI and thallium-201 in tumor cell lines. J Nucl Med. 1996;37(9):1551–6. 9. Del Vecchio S, Salvatore M. 99mTc-MIBI in the evaluation of breast cancer biology. Eur J Nucl Med Mol Imaging. 2004;31(1):S88–96. 10. Hruska CB, Scott CG, Conners AL, Whaley DH, Rhodes DJ, Carter RE, O'Connor MK, Hunt KN, Brandt KR, Vachon CM. Background parenchymal uptake on molecular breast imaging as a breast cancer risk factor: a case- control study. Breast Cancer Res. 2016;18(1):42. 11. Conners AL, Maxwell RW, Tortorelli CL, Hruska CB, Rhodes DJ, Boughey JC, Berg WA. Gamma camera breast imaging lexicon. AJR Am J Roentgenol. 2012;199(6):W767–74. 12. Conners AL, Hruska CB, Tortorelli CL, Maxwell RW, Rhodes DJ, Boughey JC, Berg WA. Lexicon for standardized interpretation of gamma camera molecular breast imaging: observer agreement and diagnostic accuracy. Eur J Nucl Med Mol Imaging. 2012;39(6):971–82. 13. Hruska CB, Rhodes DJ, Conners AL, Jones KN, Carter RE, Lingineni RK, Vachon CM. Background parenchymal uptake during molecular breast imaging and associated clinical factors. Am J Roentgenol. 2015;204(3):W363–70. 14. D'Orsi CJ, Bassett LW, Berg WA. Breast imaging reporting and data system, BI-RADS: mammography. 4th ed. American College of Radiology: Reston; 2003. 15. Byng JW, Boyd NF, Fishell E, Jong RA, Yaffe MJ. Automated analysis of mammographic densities. Phys Med Biol. 1996;41(5):909. 16. Vachon CM, Pankratz VS, Scott CG, Maloney SD, Ghosh K, Brandt KR, Milanese T, Carston MJ, Sellers TA. Longitudinal trends in mammographic percent density and breast cancer risk. Cancer Epidemiol Biomark Prev. 2007;16(5):921–8. 17. Campeau RJ, Kronemer KA, Sutherland CM. Concordant uptake of Tc-99m sestamibi and Tl-201 in unsuspected breast tumor. Clin Nucl Med. 1992; 17(12):936–7. 18. Rhodes DJ, Hruska CB, Phillips SW, Whaley DH, O'Connor MK. Dedicated dual-head gamma imaging for breast cancer screening in women with mammographically dense breasts. Radiology. 2011;258(1):106–18. 19. Hruska CB, Conners AL, Vachon CM, O’Connor MK, Shuster LT, Bartley AC, Rhodes DJ. Effect of menstrual cycle phase on background parenchymal uptake at molecular breast imaging. Acad Radiol. 2015;22(9):1147–56.
Breast Cancer Research – Springer Journals
Published: Jun 5, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
All the latest content is available, no embargo periods.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud