Tumour identification matters

Tissue microarray

There are many different variants of each type of cancer, which cannot be differentiated with current methods but where the differences can still be a matter of life and death.“We must improve our ability to identify the special characteristics of individual tumours which show the likely development of the disease and how different treatments could work”, says Karin Jirström.

Karin Jirström is a Professor of Pathology and wants to develop tools to distinguish different variants of colon and ovarian cancer. Which variants have a good prognosis and which do not? Which require harsh treatment and in which cases could new and expensive cancer drugs produce a good effect? These questions are important both for patients and for the health service. However, we still know too little to provide any good answers.

“What doctors can do is to study the tumour tissue and see where the tumour is located and whether the cancer has spread. However, these are crude methods that are insufficient to make good diagnoses and prognoses”, says Karin Jirström.
Her research involves finding proteins that could function as biomarkers to distinguish between the different variants of a certain type of cancer. When she succeeds, the difference is clearly visible.

“On this image we have a sample from an ovarian tumour with a large amount of a protein called Dach2. Some of the cell nuclei are almost black. In the light sample, on the other hand, there is almost no Dach2”, she says.

Dach2 is one of the proteins with which the research group is working. It appears to signal different variants of ovarian cancer: tumours with a lot of Dach2 are malignant and patients have a poor prognosis, whereas tumours without Dach2 are benign. In order to study the proteins, the researchers take extremely thin slices from samples of diseased and healthy tissue which have been kept enclosed in paraffin blocks. Before the thin slice of tissue is placed under the microscope, it is treated with an analysis fluid which reacts with the protein they are looking for.

The proteins being studied are taken from the ongoing project to map the human proteins. This is a huge project being carried out at KTH and Uppsala University, which will lead to the creation of an ‘atlas’ of all the 20 000–22 000 proteins in the human body. For each protein, information is recorded including where in the body it can be found and whether it is found in tumours from one of the most common types of cancer.

The work on the protein atlas has now almost reached the halfway point. New batches of proteins are published on the Internet at regular intervals and all the information is freely available. It is hoped that the new findings will lead to new ways of diagnosing and treating diseases.

The Lund group led by Karin Jirström is one of those which has done the most work to make information from the protein atlas applicable in healthcare.

“We are searching the atlas for proteins that are found in a certain type of cancer in either very large or very small quantities. Such proteins could serve as biomarkers and distinguish tumours that should have different treatments”, she explains.
Once the researchers have found such a protein, they turn to their extensive bank of tissue samples collected from tumours removed from patients. Each sample includes information about how the patient was treated and the outcome for him or her. By testing their proteins on the tissue samples, the researchers can see whether they really do distinguish between, for example, aggressive and more benign tumours as hoped.

The researchers in Lund have already found 15 interesting proteins and have taken out the same number of patents. Three of the proteins will be tested in cancer care in the next few years.
“One of our proteins is associated with colorectal cancer, the third most common cancer in Sweden. It can be used as a pointer by the doctors deciding whether a patient should have chemotherapy after a cancer operation. We have shown that if our protein is present in a tumour, the cancer is so aggressive that it should definitely be treated with chemotherapy”, says Karin Jirström.

She thinks it is advantageous to start with studies on human tissue and only after that study how the proteins behave in cell cultures and laboratory mice. Many researchers work the other way round, but the difference between mice and humans sometimes make it impossible to apply lab results in a human context.

As an example, Karin Jirström tells us about the contact she has had with a group of American researchers. The group had studied a protein that sped up cancer growth in laboratory mice. By rendering the protein harmless, they hoped to be able to cure cancer – but the Lund researchers’ studies produced different results.

“When we studied the protein in our human tissue samples we found that it was indeed also present in human cancer. However, it was only found in very benign tumours and could therefore not form the basis for any useful drug. This example shows how important it is to study human diseases in human samples”, she says.


FOOTNOTE: Genes and the proteins originating from the genes often have significant names. “Dach2”, full name “Dachshund 2”, was first found in fruit flies. The gene’s function is to give very short legs, hence the name.