PrOspEctive Metastases Imaging (POEM)

Colon cancer is a major cause of death among different cancer types. The tumor in the colon can often be treated with curative intent (by surgery, possibly in combination with medication (chemotherapy)). However, in the vast majority of patients who die from colon cancer, metastases are the real cause of death. This usually involves liver metastases. These metastases can be surgically removed and patients are often treated peri-operatively with chemotherapy or other medication although this treatment is not always successful. The existing method, eg a clinical risk score, that is being used to predict which patient will benefit from a specific treatment of liver metastases is inaccurate.

Over the past fifteen years, the investigators of the Translational Cancer Research Unit have conducted research that explains, at least partly, why not all liver metastases from colon cancer behave in the same way. These differences can be observed by histopathological examination. Surprisingly, liver metastases behave in two very different ways, or growth patterns, and this information is currently not being used in clinical practice. However, the biological differences in the behavior of the cancer cells between these two growth patterns are pronounced. In one growth pattern - "desmoplastic" - the cancer cells behave as in the primary colon cancer itself: they form groups and create new supportive tissue - or "desmoplasia" - and blood vessels - or "angiogenesis". Inflammation is often present in this type of metastases. In the other growth pattern - 'replacement' - the cancer cells behave in a special way: they take the place of, or 'replace', the normal cells of the liver, the hepatocytes, and use the available supporting tissue and blood vessels of the host. There is no formation of new blood vessels and there is often little or no inflammation. When more than one growth pattern is present in a liver metastasis, the relative contribution of each growth pattern is estimated (percentage).

These different growth patterns contain important information on the success of the treatment of the individual patient. A large retrospective study of 732 patients, where liver metastases from colon cancer were surgically removed, identified a group of patients (19%) with an exclusive desmoplastic growth pattern in all sections of the resected liver metastasis. This subgroup is associated with good survival and progression-free survival. In contrast, patients with any amount of replacement growth, have a significantly shorter overall survival. However, large retrospective studies have shown that these patients with a 'replacement' growth pattern have a survival advantage of adjuvant treatment, i.e. after surgery for liver metastasis, often a combination of chemotherapy and a treatment that inhibits angiogenesis (for example bevacizumab) is given. This is in contrast to patients with the "desmoplastic" growth pattern, the intrinsically more favorable pattern, where this treatment has no survival benefit and is therefore probably not necessary after surgery. These patients may be the so-called "oligometastatic" patients who may benefit from repeated local targeted therapies. The prognostic and predictive value of the growth patterns has been demonstrated not only in patients with colorectal cancer but also in patients with uveal melanoma, skin melanoma and breast carcinoma.

The concept of tumor growth patterns is innovative (and still underexposed in the scientific literature) and offers important perspectives for clinical application. Anti-angiogenic treatment as well as immunotherapy achieved poor results in clinical studies of metastatic colon cancer. The investigators are convinced that a better selection of patients based on characteristics of the liver metastasis, and not only based on characteristics of the primary colon tumor, could lead to success in clinical studies.

Thus, the growth patterns have an important clinical value and provide the treating physician with yet unknown information about the behavior of a liver metastasis in an individual patient. The limitation, of course, is that the growth pattern can only be determined after hepatic surgery. It would be much more useful to be able to identify the growth patterns by means of medical imaging, allowing the treatment to be tailored to the patient even before surgery. (for example; is there an additional benefit of starting chemotherapy prior to surgery?) Therefore, the aim of the current project is to ascertain whether the investigators can identify the growth patterns of these tumors by means of imaging with CT (computed tomography) or MRI (nuclear spin tomography). Some published studies suggest that medical imaging can capture the histopathological growth patterns, but these are all retrospective studies with a small number of patients and incomplete pathological sampling of the liver metastases. One study in 2007 correlated the early peri-lesional gadolinium enhancement of the liver metastasis with the presence of the desmoplastic reaction with vascular proliferation and inflammation. More recently, a study of Cheng and colleagues claimed that the prediction of the growth patterns would be possible with radiomics based on CT images.

By identifying the growth patterns in a non-invasive manner, by means of medical imaging, the investigators will be able to use the growth patterns in prospective clinical studies to truly test the predictive power.

This project concerns liver metastases of primary colorectal carcinoma as well as primary breast carcinoma. The research will therefore mainly, but not only, help patients with liver metastases of colorectal cancer and breast cancer when a translation of this project from the pre-clinical phase (explorative study) to the clinical phase (randomized studies with patients) is done.