Umeå University's logo

Background: In 2020, more than 2,260,000 women were diagnosed with breast cancer. Most patients are cured with surgery and adjuvant treatment, but despite that, approximately 700,000 women die of the disease every year. The historical focus on breast cancer progression has been on the malignant epithelial cell. However, cancer cells do not grow in isolation. In recent years, the importance of the tumour microenvironment in cancer progression has been highlighted. Perlecan and type IV collagen are basement membrane (BM) proteins in the normal mammary gland, and type I collagen is the main fibrillar collagen in the interstitial extracellular matrix (ECM). In cancer development, perlecan and type IV collagen have multifunctional roles and when degraded from the BM, bioactive substances and other fragments are released in the circulation. Significant ECM changes also occur that lead to an accumulation of fibrillar collagens. Given their abundance in the ECM; perlecan, type IV and type I collagen are of interest for breast cancer progression and may be of importance as new biomarkers to monitor disease, predict patient outcome and the treatment effect.

Aim: In this thesis, the protein and mRNA expression of perlecan, type IV and I collagen in breast cancer tissue is studied. The aim is to characterize the expression pattern of these proteins in breast cancer tissue and to see whether there is a correlation to known prognostic biomarkers and to the patient prognosis. Moreover, to evaluate circulating perlecan and type IV collagen as diagnostic and prognostic biomarkers in breast cancer patients.

Methods: In this thesis project, eight different patient cohorts were used. In freshly frozen normal breast and breast cancer tissue, perlecan protein expression was visualized using immunofluorescence. Type IV and I collagen protein expression were studied with immunohistochemistry in formalin-fixed, paraffin embedded primary breast cancer tissue, and type IV collagen in metastatic breast cancer tissue. For gene expression analysis, mRNA and clinicopathological data were extracted from the Cancer Genome Atlas and cBioportal database. Circulating plasma levels of perlecan were analysed in breast cancer patients and controls, circulating levels of CA15-3 and type IV collagen in patients with primary and metastatic breast cancer as well as controls. Perlecan and type IV collagen were measured with ELISA assays, and CA15-3 were using an electrochemiluminescence immunoassay.

Results: In breast cancer tissue, perlecan and type IV collagen protein expression in the epithelial BM was fragmented or completely lost, and perlecan and type IV collagen was expressed to varying extent in the tumour stroma. The mRNA analysis confirmed that type IV collagen mRNA was expressed in primary breast cancer tissue and highly expressed in metastatic tissue. Type I collagen was mostly highly expressed in the tumour stroma. Low type I collagen protein and mRNA expression correlated with biomarkers for aggressive breast cancer, but no effect on survival could be seen. Among patients receiving chemotherapy, low stromal type I collagen protein expression was associated with better survival compared to high expression, even after adjusting for other relevant factors. There was no correlation of perlecan or type IV collagen protein expression to clinically used prognostic biomarkers, but an oestrogen receptor dependent correlation between mRNA expression of perlecan and several matrix-degrading enzymes were found. Survival analysis showed that high stromal type IV collagen protein and mRNA expression in the primary tumour was significantly associated with a poorer survival, and high protein expression with a risk of developing distant metastasis. Metastatic breast cancer patients had higher levels of circulating type IV collagen compared to healthy controls and patients with primary breast cancer. High circulating type IV collagen levels correlated with poorer survival in metastatic breast cancer patients, and was superior to CA15-3 at detecting metastatic breast cancer.

Conclusions: The protein expression pattern of perlecan, type IV collagen and type I collagen become abnormal during breast cancer development. Stromal type IV collagen protein and mRNA in the primary tumour correlates to poorer prognosis, most likely due to a higher risk of developing metastatic disease. Circulating type IV collagen can function as a biomarker for detecting metastatic disease in patients with primary breast cancer and is prognostic in patients with metastatic breast cancer. Low stromal type I collagen is a marker for an aggressive breast cancer disease and can predict chemotherapy response.

Background: Metastatic colorectal cancer (mCRC) and metastatic breast cancer (mBC) are two leading causes of cancer-related mortality worldwide. Early detection of metastatic disease is critical, and sensitive, easily accessed and cost-effective biomarkers that can diagnose mBC and mCRC at an early stage would have high clinical value. The best circulating markers: CEA and CA 15-3 are suboptimal, but a combination with other proteins can improve their potential to detect metastatic disease. One potential source of new biomarkers is the tumor stroma, including the extracellular matrix (ECM), vasculature, and stromal cells like immune cells and fibroblasts. Both the tumor cell and stromal compartment are vital for cancer progression. A combination of biomarkers from both compartments could likely best reflect the heterogeneous nature of the metastatic disease. Stromal type IV collagen (COL IV) is the main constituent of the basement membrane of healthy tissues. COL IV is upregulated with some cancers, including colorectal liver metastases (CLM), and is considered a potential biomarker for CLM. The origin of elevated levels of COL IV in CLM is not known but may result from both increased ECM production and ECM degradation associated with cancer-related tumor stroma remodeling. 

Aims: In this thesis, COL IV and its potential to be used as a biomarker for mCRC and mBC is studied, with a specific emphasis on liver metastases. The aims are to compare levels of circulating COL IV (cCOL IV) in mCRC and mBC patients with controls and evaluate its diagnostic and prognostic value; to evaluate the combination of cCOL IV with other proteins; to determine the cellular origin of COL IV in CLM and study COL IV expression in cell lines; to study the expression of COL IV degrading proteases in CLM and to evaluate tissue expression of COL IV in bone and liver metastases from BC patients. 

Methods: Plasma levels of cCOL IV, CA 15-3, CEA, and other cancer-related proteins were analyzed with ELISA, ECLIA and Multiplex assay in mCRC and mBC patients, healthy controls and patients with primary CRC or BC as controls. The cellular origin of COL IV expression in CLM was examined with in situ hybridization, and the expression of COL IV in mBC tissue and COL IV degrading proteases (MMP -2, -7, -9 and -13) in CLMs were studied with immunohistochemistry. COL IV expression in CRC and fibroblast cell lines was analyzed with immunofluorescence.

Results: cCOL IV is elevated in mBC patients and correlates with poor survival. The combination of cCOL IV with CA 15-3 and cCOL IV alone are superior to CA 15-3 at detecting mBC. COL IV is highly expressed in the tissue of liver- and bone BC metastases. Circulating COL IV, CEA, OPN, CYFRA 21-1, IL-8, HGF, and MIF are elevated, and TRAIL is lower in mCRC patients compared with controls. COL IV, CEA, OPN, CYFRA 21-1, and IL-8 were higher, and TRAIL was lower in mCRC patients with liver metastases compared to patients with extrahepatic disease. Circulating CEA, OPN, and HGF are very good, and cCOL IV is acceptable at distinguishing mCRC patients from patients with primary CRC. The combination of OPN + CEA is superior to CEA alone at detecting mCRC. High HGF and cCOL IV (one cohort) in mCRC patients correlate to poor prognosis. cCOL IV is elevated in CLM patients compared to healthy controls and is very good at discriminating between healthy controls and CLM patients. COL IV is expressed in CLM by cancer-associated fibroblasts, and COL IV degrading proteases are expressed primarily by stromal cells in CLM. COL IV is expressed by fibroblasts, not tumor cells, in vitro. 

Conclusion: cCOL IV is a promising tumor marker of metastatic BC and CRC and circulating HGF and OPN are potential biomarkers of mCRC. Our results show that the metastatic site can impact the circulating levels of numerous cancer-related proteins, which aligns with our hypothesis that combining biomarkers instead of using one might be best for detecting metastatic cancer through blood analysis. COL IV is expressed by stromal cells, not tumor cells, in CLM tissue and in vitro.