Mammographically dense human breast tissue stimulates MCF10DCIS.com progression to invasive lesions and metastasis

Abstract

Background: High mammographic density (HMD) not only confers a significantly increased risk of breast cancer (BC) but also is associated with BCs of more advanced stages. However, it is unclear whether BC progression and metastasis are stimulated by HMD. We investigated whether patient-derived HMD breast tissue could stimulate the progression of MCF10DCIS.com cells compared with patient-matched low mammographic density (LMD) tissue.

Methods: Sterile breast specimens were obtained immediately after prophylactic mastectomy from high-risk women (n = 10). HMD and LMD regions of each specimen were resected under radiological guidance. Human MCF10DCIS.com cells, a model of ductal carcinoma in situ (DCIS), were implanted into silicone biochambers in the groins of severe combined immunodeficiency mice, either alone or with matched LMD or HMD tissue (1:1), and maintained for 6 weeks. We assessed biochamber weight as a measure of primary tumour growth, histological grade of the biochamber material, circulating tumour cells and metastatic burden by luciferase and histology. All statistical tests were two-sided.

Results: HMD breast tissue led to increased primary tumour take, increased biochamber weight and increased proportions of high-grade DCIS and grade 3 invasive BCs compared with LMD. This correlated with an increased metastatic burden in the mice co-implanted with HMD tissue.

Conclusions: Our study is the first to explore the direct effect of HMD and LMD human breast tissue on the progression and dissemination of BC cells in vivo. The results suggest that HMD status should be a consideration in decision-making for management of patients with DCIS lesions.

Keywords: Breast cancer; MCF10DCIS.com; Mammographic density; Murine biochamber.

Fig. 1

Quantitative analyses of chamber explant weights. a Coloured bar graphs show patient-matched DCIS-only (blue), DCIS + HMD (red) and DCIS + LMD (green) comparisons of chamber explant weights for all ten women. b Average values for each woman are shown as scatterplots. HMD High mammographic density, LMD Low mammographic density, DCIS MCF10DCIS.com cells. **P = 0.002. All bar and scatterplot graphs represent mean + SEM

Fig. 2

Quantitative analysis of the histopathological results of chamber material. a–d Representative photomicrographs of the various numerical categories used to score human breast histopathology observed in biochambers. a Category 0, the absence of DCIS and cancer and the presence of fatty normal breast tissue. b Category 0, the absence of DCIS and cancer and the presence of dense normal breast tissue. c Category 1, high-grade DCIS only. d Category 2-3, grade 3 invasive carcinoma. e The numerical histological categories 0–3 included a breakdown of the presence of high-grade DCIS alone or along with a certain percentage of grade 3 IDC. The mean histological category for the chamber explants for each of the ten women according to the type of input material. f Average values for each woman for the histological category of the chamber material. HMD High mammographic density, LMD Low mammographic density, DCIS MCF10DCIS.com cells, IDC Invasive ductal carcinoma. *P = 0.004; **P = 0.002. All bar and scatterplot graphs represent mean ± SEM

Fig. 3

Analyses of chamber explant luciferase signalling. a Luciferase signal measured in the chambers at explant (×10⁸ photons per second) for each of the ten women. b Average values for each woman are shown as scatterplots after removal of a significant outlier (patient 6). HMD High mammographic density, LMD Low mammographic density, DCIS MCF10DCIS.com cells. *P = 0.03; **P = 0.005. All bar and scatterplot graphs represent mean ± SEM

Fig. 4

Comparison of chamber ‘take’ rates representing the viability of input MCF10DCIS.com cells. The average number of positive biochambers (contained a histological category of at least 1: DCIS material only) as a percentage of the total number of chambers implanted. HMD High mammographic density, LMD Low mammographic density, DCIS MCF10DCIS.com cells. **P = 0.008. Error bar indicates SEM

Fig. 5

Representative images of tumour metastasis obtained by imaging and histology. a1 and a2 Images from the same mouse lymph node. b1–b3 Images illustrate the lymph node of a different mouse. a1 The white arrow indicates a luciferase (luc) signal in a mouse lymph node. a2 Haematoxylin and eosin staining of the luc-positive lymph node from a1. b1 The white arrow indicates a luc signal in a mouse lymph node. b2 Human-specific CK5 staining of the luc-positive lymph node from b1; CK5-positive cells are stained brown. b3 Human-specific Ku70 staining of the luc-positive lymph node from b1. Ku70-positive cells are stained yellow. Scale bar = 10 μm

Fig. 6

Quantitative analyses of metastases in terms of metastasis-positive organ numbers and luciferase signalling. a Coloured bar graphs of the mean value of all four mice for each group for mean metastasis + organ numbers. b Averaged values of the mean number of metastasis-positive organs for each woman according to the type of input material. c Coloured bar graphs of the mean level of luciferase in metastasis-positive organs (×10⁵ photons/second). d Averaged values of the mean luciferase load in metastasis for each woman. HMD High mammographic density, LMD Low mammographic density, DCIS MCF10DCIS.com cells. *P = 0.02; **P = 0.008. All bar and scatterplot graphs represent mean ± SEM; triangles indicate data value of 0

Fig. 7

Heat map of metastasis profile for each woman. The degree of metastasis is depicted using filled blue squares to denote metastasis-positive organs, and the level of metastatic burden is shown using the colour code shown. Patients 1–10 are shown in rows. For each woman, four mice were implanted with DCIS alone, four with DICS + HMD and four with DCIS + LMD. The grouped columns represent the types of cells implanted, and the sub-columns represent the site of metastatic burden. The numbers in boldface type at the beginning of each row represent the chamber explant weight to show the correlation between chamber weight and metastasis. Inside each bar, the luciferase levels in photons per second are indicated

Fig. 8

CTC analyses. a Representative photomicrographs under bright field of CTCs isolated and cultured blood from mice blood. b Corresponding fluorescence image to demonstrate RFP tagging in CTCs. c Representative photomicrographs of CTCs after crystal violet staining. d–f As negative controls, blood taken from naïve mice was also processed and cultured, and images were obtained to show the absence of CTCs. g Coloured bar graphs of the mean value of all four mice for each group for CTC numbers. h Averaged values of the mean number of CTCs for each woman according to the type of input material. CTC Circulating tumour cell, HMD High mammographic density, LMD Low mammographic density, DCIS MCF10DCIS.com cells, RFP Red fluorescent protein. All bar and scatterplot graphs represent mean ± SEM. Scale bar = 50 μm