Denosumab as an immune modulator in HER2-negative early breast cancer: results of the window-of-opportunity D-BIOMARK clinical trial

Abstract

Background: The RANK pathway has been extensively investigated for its role in bone resorption; however, its significance extends beyond bone metabolism. Preclinical models suggest that inhibition of RANK signaling can prevent mammary tumor development by reducing proliferation and tumor cell survival. Additionally, both preclinical and clinical data support the ability of RANK pathway inhibitors to enhance the anti-tumor immune response.

Methods: D-BIOMARK is a prospective, randomized window-of-opportunity clinical trial assessing the biological effects of denosumab, a monoclonal antibody against RANKL, in patients with HER2-negative early breast cancer. The study aims to assess denosumab's impact on breast tumor cell proliferation, apoptosis, and its potential to influence the tumor immune microenvironment. A total of 60 patients were enrolled and randomized 2:1 to receive two doses of single agent denosumab (120 mg one week apart) before surgery or to the control arm (no treatment). Fifty-eight patients were evaluated, 27 pre-menopausal and 31 post-menopausal women, 48 with luminal tumors and 10 with triple negative breast cancer. Paired tumor samples were collected to compare baseline (core biopsy) and surgical (surgical specimen) time points, as well as serum samples at both time points.

Results: Denosumab demonstrated its ability to reduce serum free RANKL levels (experimental p < 0.001, control p = 0.270). However, a reduction in tumor cell proliferation or cell survival was not observed. A denosumab-driven increase in tumor infiltrating lymphocytes (TILs) was observed (experimental p = 0.001, control p = 0.060), particularly in the luminal B-like population (experimental p = 0.012, control p = 0.070) and a similar trend in the TNBC group (experimental p = 0.079, control p = 0.237). Denosumab led to increased TILs in both pre-menopausal (experimental p = 0.048, control p = 0.639) and post-menopausal (experimental p = 0.041, control p = 0.062) women with luminal tumors. RANK protein expression in tumor and stroma was associated with markers of tumor aggressiveness but an increase in TILs was observed in the experimental arm, irrespectively of RANK and RANKL expression in tumor or stromal cells.

Conclusions: The D-BIOMARK trial suggests a potential role for denosumab as an immune-enhancing agent in early HER2-negative breast cancer. Although preoperative denosumab did not reduce tumor proliferation or increased apoptosis, it led to an increase in TILs, particularly in luminal B-like tumors. These findings underscore the importance of further investigation into the multifaceted aspects of the RANK pathway. Trial registration EudraCT number: 2016-002678-11 registered on June 15, 2018.

Clinicaltrials: gov identifier: NCT03691311, retrospectively registered on September 04, 2018.

Keywords: Breast cancer; Denosumab; HER2-negative; Immune enhancer; RANK; RANKL; TILs; Tumor cell proliferation; Tumor cell survival.

Fig. 1

D-BIOMARK trial design. Diagram showing the steps to be followed by a patient from the time the Informed Consent Form (ICF) is signed until the end of the study

Fig. 2

CONSORT Flow diagram. The CONSORT flow chart illustrates the flow of participants throughout the D-BIOMARK study. A total of 60 patients were initially enrolled, with 58 patients evaluated in the final analysis, 37 patients in the experimental arm and 21 patients in the control arm. The analysis was performed according to protocol

Fig. 3

Serum biomarkers: free RANKL (sRANKL), Osteoprotegerin (OPG), Tartrate-resistant acid phosphatase 5b (TRACP-5b) and Calcium. Levels of sRANKL, OPG, TRACP-5b, detected by ELISA, and Calcium in serum from patients collected at the time of biopsy (Serum A) and surgery (Serum B) in the control and experimental arms. p value T test for comparison between experimental and treatment arm is shown in the upper left corner and p value T test for paired samples is shown for each treatment group. N indicates the number of samples analyzed. Note that Denosumab was associated with reduction in the levels of free RANKL and serum calcium, OPG shows a tendency to increase, while the levels of Trap5b did not change

Fig. 4

Impact of Denosumab on Tumor Cell Proliferation (Ki67) and apoptosis (cleaved caspase-3). Percentage of Ki-67 + tumor cells (A) and H-score of cleaved caspase-3 in tumor cells (B) at the diagnostic biopsy (biopsy) and in the surgical specimen (surgery) in the control and experimental arm. p value t test for comparison between experimental and treatment arm is shown in the upper left corner and p value t test for paired samples is shown for each treatment group. N indicates the number of samples analyzed. C and D illustrate representative immunohistochemistry staining images from the control and experimental groups for Ki67 and cleaved caspase-3, respectively, P denotes the selected patient. Note that the percentage of Ki67-positive cells in most cases was intermediate (around 20–30%), with a slight, non-clinically relevant increase observed in the surgical specimens. Regarding the apoptosis marker cleaved caspase-3, it was scarcely positive in a few cases, with no changes observed in either group. Comparable mean levels of Ki67 and cleaved caspase-3 were observed between both groups

Fig. 5

Impact of Denosumab on stromal Tumor Infiltrating Lymphocytes (TILs). A: Percentage of stromal TILs at the diagnostic biopsy (biopsy) and in the surgical specimen (surgery) in the control and experimental arm. p value T test for comparison between experimental and treatment arm is shown in the upper left corner and p value t-test for paired samples is shown for each treatment group. N indicates the number of samples analyzed. B: Representatives images of H&E staining depicting changes in TILs between biopsy and surgery in patients from the control and experimental arm. Black arrows indicate regions with TILs, P denotes the selected patient. Note that only the experimental group exhibited a statistically significant increase in the percentage of TILs

Fig. 6

Correlation analysis between RANK/RANKL expression at baseline in tumor or stroma and clinicopathological parameters. A: Representative image of RANK expression in tumor cells; B: Representative image of RANK expression in stromal cells; C: Representative image of RANKL expression in tumor cells; and D: Representative image of RANKL expression in stromal cells. The molecular surrogate subtype is indicated in parentheses, P denotes the selected patient. E: Correlation analysis between RANK/RANKL expression in tumor with Ki67, histological grade, ER and TILs using Pearson’s correlation coefficient. F: Correlation analysis between RANK/RANKL expression in stroma with Ki67, histological grade, ER and TILs using Pearson’s correlation coefficient. Note that RANK expression in tumor cells positively correlates with Ki67, histological grade, and negatively correlates with estrogen receptor expression. Additionally, RANK expression in stroma is associated with high Ki67 levels. No significant correlation was found with RANKL expression