< Back to previous page

Project

Linking patient adiposity to tumour microenvironment and therapy response in breast cancer: towards tailored treatment strategies in obesity

This thesis explores innovative translational research approaches in breast cancer, aiming to increase biological understanding of the disease and thereby help advance precision medicine. It does so in two groups of patients with a high clinical unmet need. First, it investigates patients with early breast cancer in the context of adiposity (body fat accumulation). Secondly, it investigates metastatic breast cancer, which is the ultimate reason for death of cancer but is poorly investigated due to a lack of sample availability.

Adiposity is associated with a higher risk of postmenopausal estrogen receptor (ER)-positive breast cancer and poorer breast cancer-specific outcomes, but patients are largely treated independently of their body composition. Insights into the biological link between the two diseases still largely come from preclinical research. As the effects of adiposity on the tumor micro-environment likely play a role, it is crucial to assess these interactions in patients on a cellular level.

In Chapter 3, we present preliminary results of the FATLAS trial, in which adiposity measures, tumor samples and macroscopically normal breast samples were prospectively collected from patients with stage I-III ER-positive HER2-negative invasive breast cancer of no special type (n=64). We show the correlations between measures of adiposity both histologically (adipocyte measurements, crown-like structures) and systemically (leptin levels, visceral fat percentage,…). Next, we performed single nuclei RNA-sequencing in the tumor samples from 11 lean, 11 overweight and 11 obese patients matched for menopausal status and tumor-infiltrating lymphocyte levels. Globally, cell composition did not differ between obese and lean individuals. A low number of cancer-associated adipocytes was identified. Differential gene expression analysis between obese and lean patients then showed diverse obesity-associated effects on the transcriptomic profile of different cell populations, particularly T cells, B cells, mast cells, and adipocytes. When grouping patients based on adipocyte sizes or the presence of local inflammation instead, we observed different associations with cell type-specific transcriptomic profile compared to those identified for obese versus lean. While these results highlight the potential of our analyses, ongoing efforts will be undertaken to increase the quality control measures on our dataset and answer additional questions on differential breast cancer biology according to adiposity.

In Chapter 4.1, we present the shared experience of 14 rapid autopsy programs in oncology worldwide to highlight the research opportunities and challenges in implementation of such programs. Different inclusion strategies are described, including focused approaches to maximize patient participation without causing psychological distress. In terms of study conduct, eight programs operate 24 hours/7 days, resulting in a lower median post-mortem interval (4 hours) compared to those operating during (extended) working hours (9 hours). Most programs (n=10) succeed in collecting all samples within a median of 12 hours after death. A large number of samples is obtained during each autopsy (median of all medians: 58 samples/patient), representing both liquid, tumor and non-tumor tissues. Unique biological insights derived from these samples include metastatic progression, treatment resistance, disease heterogeneity, interactions with the tumor micro-environment and representability of liquid biopsies. Additionally, tumor patient derived xenograft (PDX) or organoid (PDO) models can be established from the samples, allowing for drug discovery and treatment sensitivity assays. Through this work, we hope to increase visibility and transparency of post-mortem tissue donation programs, foster collaborations as well as to encourage the creation of new programs.

In Chapter 4.2, we show how the breast cancer post-mortem tissue donation program UPTIDER enabled the sampling of a median of 31 (range: 5-90) metastases and 5-8 liquids per patient (n=20). We demonstrate vast macroscopic underestimation of disease extent in lobular breast cancer and mechanistic insights into the disruption of the blood-brain barrier in leptomeningeal metastasis. We also highlight important intra-patient heterogeneity of the acknowledged biomarkers estrogen receptor, progesterone receptor, KI67 and tumor-infiltrating lymphocytes. We successfully generated ex vivo models from both tissue and liquid biopsies from patients with distinct histological subtypes. Lastly, we show how cooling can mitigate the mild impact of increasing time after death on transcriptomic and protein expression in tumor tissue samples. We anticipate these and future findings of UPTIDER to elucidate mechanisms of disease progression and treatment resistance and to provide tools for the exploration of precision medicine strategies.

And finally, in Chapter 4.3, we additionally highlight intra-patient heterogeneity of HER2-expression. Anti-HER2 antibody-drug conjugates (ADCs) have shown important efficacy in HER2-low (HER2-1+ or HER2-2+ with negative in situ hybridisation [ISH] according to the current scoring guidelines) metastatic breast cancer. Criteria for receiving ADCs are based on a single assay on the primary tumor or a small metastatic biopsy. We assessed the intra-patient inter-metastasis heterogeneity of HER2-low status in HER2-negative mBC. We included samples of 10 patients (7 ER-positive and 3 ER-negative) donated in the context of our post-mortem tissue donation program UPTIDER. Excisional post-mortem biopsies of 257 metastases and 8 breast tumors underwent central HER2 immunohistochemistry (IHC), alongside 41 pre-mortem primary or metastatic samples. They were classified as HER2-zero, HER2-low (HER2-1+ or HER2-2+, ISH negative) or HER2-positive (HER2-3+ or HER2-2+, ISH-positive) following ASCO/CAP guidelines 2018. HER2-zero was further subdivided into HER2-undetected (no staining) and HER2-ultralow (faint staining in ≤10% of tumor cells). In 8/10 patients, HER2-low and HER2-zero metastases co-existed, with the proportion of HER2-low lesions ranging from 5% to 89%. A total of 32% of metastases currently classified as HER2-zero were HER2-ultralow. Intra-organ inter-metastasis heterogeneity of HER2-scores was observed in the liver in 3/6 patients. Patients with primary ER-positive disease had a higher proportion of HER2-low metastases as compared to ER-negative disease (46% versus 8%, respectively). At the metastasis level, higher percentages of ER-expressing cells were observed in HER2-low or -ultralow as compared to HER2-undetected metastases. In conclusion, we show important intra-patient inter-metastasis heterogeneity of HER2-low status. This questions the validity of HER2-low in its current form as a theranostic marker.

In conclusion, this thesis uses innovative approaches (single nuclei sequencing and extensive collection of adiposity measures in Chapter 3, post-mortem tissue donation for translational research in Chapter 4) to advance our biological understanding of breast cancer in patients with a high clinical unmet need. We are hopeful our efforts will contribute to better precision medicine, and ultimately better outcomes, in these patients.

Date:27 Sep 2019 →  30 Sep 2023
Keywords:breast cancer
Disciplines:Cancer therapy
Project type:PhD project