Histologic correlates of background parenchymal uptake on molecular breast imaging

Abstract

Background

Molecular breast imaging (MBI) has shown utility as a supplement to mammography in women with dense breasts. Background parenchymal uptake (BPU), which describes the level of Tc-99 m sestamibi in fibroglandular tissue on MBI, varies among women with similar density. Higher BPU is associated with greater breast cancer risk. The goal of this work is to investigate the histological correlations of BPU that may explain this association with risk.

Methods

In healthy women with dense tissue on mammography, we prospectively biopsied regions of fibroglandular tissue exhibiting one of the extreme categories of BPU on MBI: either photopenic or marked. Tissue composition of specimens was assessed by measuring area of epithelium, stroma, and fat and counting lobules. Lobular involution status was assessed by an expert pathologist as either none, partial, or complete. Ki-67 index and estrogen receptor-alpha expression were also assessed. Histologic features of photopenic and marked specimens were compared.

Results

Biopsies were performed in 48 women (mean age 54.9 years [SD 11.1 years]), including 20 with marked BPU and 28 with photopenic BPU. Women with marked BPU were younger (mean age 49.9 vs. 58.4 years, p = 0.004), had higher body mass index (p = 0.01), and were more likely premenopausal (p = 0.005). Marked BPU specimens had a higher proportion of epithelium (14.5% vs. 2.2%, p < 0.001), lower stromal content (44.8% vs. 84.2%, p = 0.005), similar fat (18.9% vs. 12.9%, p = 0.41), and greater lobule counts (15.5 vs. 6.3, p < 0.001) compared to photopenic specimens. Complete lobular involution was less frequent in marked tissue than in photopenic tissue (10.5% vs. 76%, p < 0.001). Ki-67 index was higher in marked BPU tissues (5.4% vs. 1.6%, p = 0.006), though the difference was attenuated after adjustment for epithelial area (p = 0.26). Estrogen receptor-alpha expression did not differ between marked and photopenic groups (25.6% vs. 28.4%, p = 0.12).

Conclusions

Marked BPU on MBI is associated with greater epithelial content, less lobular involution, and higher proliferative activity—features characteristic of tissue at elevated risk for malignant transformation. These findings suggest that BPU provides functional information beyond mammographic density, supporting its role as an imaging biomarker for breast cancer risk in women with dense breasts.

Trial registration

NCT01240278 || 11/15/2010 and NCT01588834 || 05/01/2012.

Background

Molecular breast imaging (MBI), a nuclear medicine test performed with a dedicated gamma camera, has shown utility as a supplement to mammography for detecting cancer in women with dense breast tissue [1, 2]. MBI detects uptake of a radiotracer, Tc-99m sestamibi, that preferentially accumulates in cellular mitochondria and is increased with angiogenesis and blood flow, features associated with breast cancers. In addition, Tc-99m sestamibi may accumulate to varying degrees within the breast’s benign fibroglandular tissue. This feature, called “background parenchymal uptake (BPU)”, on MBI was initially regarded as a diagnostic impediment to breast cancer detection, but has since been shown to have prognostic significance among women with dense breasts [3].

While mammographic density is well-established as a breast cancer risk factor, BPU provides additional risk information beyond breast density. BPU has an interesting association with mammographic breast density: among women with dense breasts, fibroglandular tissue that appears similarly dense on mammography can show a range of BPU levels on MBI, from very little (photopenic) to very intense (marked) (Fig. 1). In a case–control analysis including over 3000 women, those with dense breasts and high BPU had 3.3-fold to 4.6-fold increased risk for breast cancer compared to women with dense breasts and low BPU, after adjusting for density-related risk [4]. With follow up of this cohort (median follow up of 7.3 years), the 5-year absolute risk of breast cancer was 4.3% for women with elevated BPU compared to 2.5% those with low BPU [5]. These findings suggest that the functional information from MBI may further stratify risk among the large population of women with dense breast tissue.

Fig. 1.

Images from four women with similar mammographic density but different background parenchymal uptake (BPU) on molecular breast imaging (MBI). Right breast mediolateral oblique views are shown for the full field digital mammography and MBI for all four categories

It is unclear why dense fibroglandular tissue exhibits varying levels of BPU across women with dense breasts and how this additional imaging feature may confer breast cancer risk. Dense tissue is heterogeneous, consisting of varying proportions of glandular epithelium and stromal components [6, 7]. Ghosh et al. examined biopsy specimens of benign breast tissue to study the relationship between breast density and lobular involution, a process of age-related physiological atrophy of breast epithelium that is linked to reduced breast cancer risk. While non-dense tissue was strongly associated with complete lobular involution, no such association was observed for dense tissue. The investigators hypothesized that during lobular involution, lobules in the epithelium are replaced either by fat, which appears non-dense on mammography, or by stroma, which can contribute variably to radiographic density [8]. Mammographic imaging is unable to differentiate between epithelial and stromal contributions to breast density; however, we hypothesized that BPU on MBI may reflect the relative proportions of epithelium and stromal tissues and indicate the tissue’s lobular involution status.

The goal of this work is to investigate BPU’s histological correlations, including tissue composition and lobular involution status, that may explain its association with breast cancer risk.

Methods

Study cohort

This research was approved by the Mayo Clinic Institutional Review Board. Written informed consent was obtained from all participants. Study recruitment and collection of specimens occurred from February 2011 to May 2014. Follow up for breast cancer events was performed through May 2025.

Categories of BPU patterns are based on the intensity of radiotracer uptake in breast fibroglandular tissue relative to uptake in subcutaneous fat: 1) photopenic—less than subcutaneous fat; 2) mild BPU—equal or slightly greater than subcutaneous fat; 3) moderate BPU—greater than mild, but less than twice as intense as subcutaneous fat; 4) marked BPU—twice as intense as subcutaneous fat [9]. Figure 1 shows the four BPU categories with their corresponding mammograms.

Healthy female volunteers, age 40 and older, with a region of mammographically dense tissue (BI-RADS categories C and D) amenable to biopsy that exhibited one of the two extreme categories on MBI, photopenic or marked, were eligible. MBIs performed within 3 months prior to enrollment were used to assess eligibility. The reproducibility of BPU categorical assignments has been studied in a previous paper [5]. Hormonal factors including menopausal status and current use of exogenous estrogen therapy were documented. Participants were classified as postmenopausal if one of the following criteria were met: ≥ 12 consecutive months of amenorrhea, follicle-stimulating hormone (FSH) level > 40 mIU/mL, surgical menopause via bilateral oophorectomy, or age ≥ 60 years. Otherwise, participants were classified as premenopausal. Subjects were excluded if they had taken estrogen receptor–modulating drugs (e.g., tamoxifen, raloxifene) or aromatase inhibitors within six months prior to biopsy, had a history of breast cancer within the past three years, or had breast implants.

Biopsy

A breast radiologist along with the study principal investigator selected a target for biopsy by reviewing the participant’s mammogram and MBI exam to identify an area of the breast with the greatest mammographic density that correlated to either photopenic or marked BPU on the MBI. Selected targets were in a region most amenable to biopsy (i.e. not near nipple, axilla or chest wall) and well-visualized as fibroglandular tissue on ultrasound. After injection of local anesthetic, the radiologist performed a 14-gauge core needle biopsy of this area using ultrasound guidance. Two cores of tissue were obtained from the biopsy site and preserved in formalin. A metallic marker clip was inserted at the biopsy site to enable subsequent confirmation of the sampled location on post-procedure mammogram. Standard post-procedure biopsy care was performed.

Histopathologic assessments

Samples were from frozen tissues that were sectioned into slides at five microns. All slides were used for pathology review; between 10–11 slides were available for each participant. Slides were stained with Hematoxylin and Eosin (H&E). Slides were scanned at 20 × magnification on the Aperio AT Turbo brightfield instrument (Leica Biosystems). Resolution of the images was 0.50 µm/pixel. The whole slide scans were 24-bit contiguous standard pyramid tiled TIFFs compressed via JPEG with a quality setting of 70. Quantitative measurements for tissue composition were conducted using Aperio ImageScope software (Leica Biosystems). Blinded to BPU group and all other participant characteristics, the area of stroma, epithelium, and fat on each slide was manually segmented by a cytotechnologist under the guidance of an anatomic pathologist.

Immunohistochemical (IHC) staining was performed for Ki-67 as an index of cellular proliferation and estrogen receptor-alpha (ERα) using standard antibodies (Ki-67: MIB-1, Dako; ERα: Clone ID5, Dako). Slides were scanned as stated above and analyzed using Aperio ImageScope. A cytotechnologist traced the epithelium regions taking care to avoid tissue folds and staining artifact. A proprietary nuclear analysis algorithm (Aperio) was applied for quantitative analysis generating a percentage of the total cells staining positive. A portion of cases were reviewed by an anatomic pathologist who verified analysis results.

Lobular involution status

Lobular involution was assessed by an expert pathologist by visual examination of H&E-stained slides, as previously described [10]. The pathologist was blinded to BPU group and all other participant characteristics such as age and menopausal status. The number of lobules in each specimen were counted and recorded. The level of involution was categorized into three groups: no involution (0% of terminal duct lobular units (TDLUs) involuted), partial involution (1% to 74% involuted), and complete involution (75% or more of TDLUs involuted) [8, 11].

Exploratory breast cancer outcome assessment

As an exploratory analysis, we investigated potential differences in long-term breast cancer outcomes between participants with photopenic and marked BPU. We conducted passive follow-up of all study participants to determine the incidence and type of any breast cancer diagnoses. Electronic medical records from our institution and linkages provided via Epic Care Everywhere were reviewed for each subject from the date of study enrollment through May 2025. Cancer diagnoses were confirmed using pathology reports, imaging records, and clinical notes. Subjects without a documented diagnosis of breast cancer by May 2025 were considered cancer-free for the purposes of this analysis.

Statistical analysis

Data are summarized using frequencies and percentages for categorical data, and either means and standard deviations (SD) or medians and interquartile ranges (IQR) or ranges for continuous data, depending on the distribution of the data. Subject demographics were compared between those with photopenic BPU and marked BPU using Chi-square or Fisher’s exact tests for categorical data, and either a t-test or Wilcoxon rank sum test for continuous data. Wilcoxon rank sum tests were used to test for differences in the proportions of stroma, epithelium, and fat between marked BPU and photopenic BPU specimens, with individual adjustments for age, BMI, and menopausal status. Wilcoxon rank sum test was also used to compare the proportions of Ki-67 and ERα positive cells. As Ki-67 and ERα positive cells were only present in epithelium, results were adjusted for total area of epithelium. The number of lobules was compared between marked BPU and photopenic BPU specimens using a t-test. Linear regression was used to assess this association after individually adjusting for age, BMI, and menopausal status. Fisher’s exact test was used to compare lobular involution status between groups. This comparison was also individually adjusted for participant age, BMI, and menopausal status using logistic regression. The Kaplan–Meier method was used to estimate rates of breast cancer diagnosis within 5 and 10 years after study participation.

Results

Study cohort

A total of 48 women were enrolled in this study, including 20 with marked BPU and 28 with photopenic BPU. Self-reported race was white in 47 of the women and one was American Indian / Alaskan Native. Table 1 gives the subject demographics for both groups of women. Women with marked BPU were younger, and more likely to be premenopausal compared to those with photopenic BPU. Additionally, women with marked BPU had higher body mass index, and were less likely to have extremely dense breasts (BI-RADS category D) compared to those with photopenic BPU. There was no evidence of a difference in current estrogen therapy use, personal history of breast cancer, or Gail model risk between marked BPU and photopenic BPU and groups.

Table 1.

Subject demographics

Characteristic	All (N = 48)	Photopenic BPU (N = 28)	Marked BPU (N = 20)	p
Age at biopsy (y)–mean ± SD (range)	54.9 ± 11.1 (42–74)	58.4 ± 12.0 (42–74)	49.9 ± 7.4 (42–67)	0.004
Body mass index–mean ± SD (range)	26.4 ± 5.4 (18.2–41.9)	24.6 ± 3.7 (18.2–37.3)	28.9 ± 6.4 (18.3–41.9)	0.01
BI-RADS density–N(%)				0.034
B	5 (10)	2 (7)	3 (15)
C	32 (67)	16 (57)	16 (80)
D	11 (23)	10 (36)	1 (5)
Menopausal status				0.005
Premenopausal	22 (46)	8 (29)	14 (70)
Postmenopausal	26 (54)	20 (71)	6 (30)
Current estrogen therapy				0.78
Yes	13 (27)	8 (29)	5 (25)
No	35 (73)	20 (71)	15 (75)
Personal history of breast cancer				0.50
Yes	2 (4)	2 (7)	0 (0)
No	46 (96)	26 (93)	20 (100)
Gail model 5-year-risk (%)–median [IQR] (range)	1.4 [1.0, 1.8] (0.5–6)	1.6 [1.1, 2.0] (0.5–5.9)	1.3 [0.8, 1.7] (0.6–6)	0.26
Gail model lifetime risk (%)–median[IQR] (range)	9.0 [7.3, 12.5] (3.5–46.8)	8.7 [6.4, 11.9] (3.5–18)	9.7 [8.0, 14.5] (4.8–46.8)	0.17

Histopathologic assessments

Table 2 shows the results of the tissue composition by BPU group. Specimens with marked BPU had a higher proportion of epithelium, lower proportion of stroma, and similar proportion of fat compared to photopenic BPU specimens.

Table 2.

Tissue composition

Component	Photopenic BPU (N = 28)		Marked BPU (N = 20)		p*
	Area (mm2)	Proportion (%)	Area (mm2)	Proportion (%)
Epithelium	0.3 (0.1, 0.6)	2.2	1.7 (0.4, 3.7)	14.5	< 0.001a
Stroma	9.2 (6.6, 14.0)	84.2	5.8 (1.2, 12.4)	44.8	0.005b
Fat	1.5 (0.4, 5.1)	12.9	1.6 (0.5, 6.3)	18.9	0.41c

^*Comparison of proportion of total area between photopenic BPU and marked BPU specimens

^ap = 0.001 after adjusting for age, p < 0.001 after adjusting for BMI, and p = 0.002 after adjusting for menopausal status

^bp = 0.013 after adjusting for age, p = 0.003 after adjusting for BMI, and p = 0.009 after adjusting for menopausal status

^cp = 0.37 after adjusting for age, p = 0.17 after adjusting for BMI, and p = 0.24 after adjusting for menopausal status

Figures 2 and 3 show examples of the MBI images, biopsy clips, and corresponding biopsy tissue that was stained from a subject with marked BPU and a subject with photopenic BPU, respectively. The specimen from marked BPU showed a greater proportion of epithelium versus stroma compared to the photopenic BPU specimen. This can be seen when comparing Figs. 2C and 3C.

Fig. 2.

Biopsy site localization and tissue specimen from a 62-year-old woman with marked BPU. A MBI image showing the biopsy site, B full-field digital mammogram confirming biopsy clip placement, and C corresponding hematoxylin and eosin (H&E) stained core biopsy specimen

Fig. 3.

Biopsy site localization and tissue specimen from a 58-year-old woman with photopenic BPU. A MBI image showing the biopsy site, B full-field digital mammogram confirming biopsy clip placement, and C corresponding hematoxylin and eosin (H&E) stained core biopsy specimen

Ki-67 index was higher in marked BPU specimens compared to photopenic (median 5.4% vs. 1.6%, p = 0.006) but did not differ between groups after adjustment for area of epithelium (p = 0.26). Additionally, ERα expression did not differ between marked and photopenic BPU groups (median 25.6% vs. 28.4%, p = 0.12).

Lobular involution status

Table 3 shows the results of the lobular involution status across by BPU group. Marked BPU specimens had more lobules on average and were less likely to exhibit complete lobular involution compared to the photopenic BPU specimens.

Table 3.

Lobular involution status

	Photopenic BPU (N = 25)*	Marked BPU (N = 19)*	p	Age-adjusted p	BMI adjusted p	Menopause adjusted p
Number of lobules	6.3 ± 4.9	15.5 ± 9.5	< 0.001	0.001	< 0.001	0.002
Lobular involution			< 0.001	< 0.001	< 0.001	< 0.001
None	0	2 (10.5)
Partial	6 (24)	15 (79)
Complete	19 (76)	2 (10.5)

^*Only specimens with at least 1 lobule identified were eligible for analysis

Breast cancer outcome assessment

In the 48 participants with median follow up of 11.5 years (range 0.9–14 years), 6 were diagnosed with breast cancer at median 5.6 years (range 1.6–9.5 years) after study participation. The median time to diagnosis and Kaplan–Meier estimated rates of breast cancer within 5 and 10 years of study participation are given in Table 4.

Table 4.

Kaplan–Meier estimated rates of breast cancer diagnosis within 5 and 10 years after study participation

	All (N = 48)	Photopenic BPU (N = 28)	Marked BPU (N = 20)
Number of participants with breast cancer diagnosis	6	3	3
Time to diagnosis (y)–median (range)	5.6 (1.6–9.5)	6.1 (5.1–8.2)	4.0 (1.6–9.5)
5-year breast cancer incidence (95% CI)	4.3% (1.1% to 16.5%)	0%	10% (2.7% to 37.2%)
10-year breast cancer incidence (95% CI)	13.2% (6.2% to 27.8%)	11.3% (3.9% to 32.7%)	15.6% (5.5% to 44.4%)

Of 20 participants with marked BPU, three had subsequent breast cancer including the following diagnoses: high grade ductal carcinoma in situ; grade 1, 0.9 cm invasive ductal carcinoma; and grade 1, 1.4 cm invasive ductal carcinoma. Of 28 participants with photopenic BPU, 3 had a subsequent breast cancer including a low-grade ductal carcinoma in situ (DCIS); a grade 1, 5 mm invasive papillary carcinoma; and a grade 2, 0.6 cm invasive ductal carcinoma.

Discussion

This study determined distinct histopathologic features associated with background parenchymal uptake (BPU) on molecular breast imaging (MBI), providing insight into the tissue characteristics that may explain its association with breast cancer risk. We found that women with marked BPU had breast tissue with significantly higher epithelial content, more lobules, and less lobular involution compared to women with photopenic BPU. These findings are consistent with the hypothesis that accumulation of Tc-99m sestamibi in areas of fibroglandular tissue exhibiting elevated BPU reflects more metabolically active, proliferative tissue, potentially contributing to an elevated risk of breast cancer.

The association between reduced lobular involution and marked BPU is particularly noteworthy. Prior studies have shown that lobular involution, the age-related regression of TDLUs, is associated with lower breast cancer risk. Our findings extend this concept by suggesting that BPU may serve as a functional imaging biomarker for the degree of involution. Specifically, marked BPU appears to correspond to tissues with preserved epithelial structures and limited involution—features indicative of ongoing cellular activity and, potentially, a tissue microenvironment more susceptible to malignant transformation.

We also observed higher Ki-67 index in marked BPU tissues, indicating that BPU may reflect greater proliferative activity. However, this difference in Ki-67 between marked and photopenic specimens was attenuated after adjusting for epithelial content, suggesting that the higher Ki-67 index was driven by greater volume of epithelium and not necessarily an increased proliferation per cell. ERα expression did not significantly differ between BPU categories, indicating that differences in metabolic activity observed on MBI are not explained by ERα status alone. Similar findings were previously reported in the comparison of dense and non-dense breast biopsy specimens by Ghosh et al. [12] regarding both Ki-67 and ERα.

To our knowledge, this is the first investigation into histologic correlates of BPU on MBI. BPU as an imaging feature has similar characteristics to background parenchymal enhancement (BPE) as observed on breast magnetic resonance imaging (MRI). Both BPU and BPE are influenced by hormonal factors and have an association with breast cancer risk [3]. However, the mechanisms of BPU and BPE differ. BPU reflects trapping of Tc-99m sestamibi within cellular mitochondria while BPE reflects a temporal change in gadolinium contrast entering the blood and extracellular spaces. A study by Sung et al. [13] found a strong correlation with microvessel density and BPE among pre-menopausal patients, however this correlation was not seen in postmenopausal patients. To date, investigations into histologic correlates of background uptake on MBI, MRI, and other modalities have been sparse and warrant further studies.

A limitation of our study was the small sample size, which did not permit multivariable analyses. While this study included long-term data with a median follow-up of over 12 years, the number of breast cancer events was small (n = 6) limiting power for comparison of cancer rates between BPU categories. Further studies with a larger cohort are necessary to examine whether the histologic features associated with elevated BPU also contribute to greater breast cancer incidence or more aggressive cancer subtypes.

Another potential limitation was that selection of an area of fibroglandular tissue for biopsy was subject to radiologists’ preference, correlation of the mammogram with MBI, and the operator dependence of ultrasound. However, the variability in biopsy target selection across radiologists and participants was mitigated by involvement of the study principal investigator for all selections and verification of the sampled area through a post-procedure mammogram with visualization of the marking clip.

Taken together, these results support the concept that BPU on MBI captures biologically meaningful variation in breast tissue composition that is not discernible on mammography. Our findings provide proof of principle that BPU may help stratify women with dense breasts into more refined risk categories. Specifically, women with dense breasts but photopenic BPU may have relatively lower-risk tissue, whereas those with dense breasts and marked BPU may have higher-risk characteristics than mammographic density alone would suggest.

Conclusion

In summary, background parenchymal uptake on molecular breast imaging reflects underlying histologic differences in epithelial content, lobular involution, and proliferative activity. These findings support the potential of BPU as a functional imaging biomarker for breast cancer risk stratification, particularly among women with dense breast tissue.

Abbreviations

MBI

Molecular breast imaging

BPU

Background parenchymal uptake

FSH

Follicle-stimulating hormone

H&E

Hematoxylin and eosin

IHC

Immunohistochemical

Erα

Estrogen receptor-alpha

TDLUs

Terminal ductal lobular units

SD

Standard deviation

IQR

Interquartile ranges

BI-RADS

Breast imaging reporting and data system

DCIS

Ductal carcinoma in situ

Author contributions

CH, KG, DR, and CMV designed the study and directed analyses. LBS performed participant follow up and was a major contributor to writing the manuscript. ALC, KNH, and RWM performed study biopsies. TNK and DWV performed histologic and pathologic assessments. KCC and CGS performed statistical analyses. All authors read and approved the final manuscript.

Funding

Mayo Clinic Center for Women’s Health.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This research was approved by the Mayo Clinic Institutional Review Board; the trial registration was NCT01240278 and NCT01588834. Written informed consent was obtained from all participants.

Consent for publication

Informed written consent was obtained from all participants.

Competing interests

CB Hruska receives royalties for technologies licensed to CMR Naviscan, a manufacturer of MBI systems.