DT Extra: Addicted to sun?

Researchers have discovered that there are many aspects of wound healing reminiscent of tumor biology. Finding the wound-response gene signature in cancers, such as breast cancer and possibly melanoma, could help physicians better determine who does and does not need adjuvant therapy.

The concept, says Howard Chang, M.D., the study's lead author, starts with the idea that wound healing is a special emergency repair response, which allows cells to break their normal rules of behavior.

"In normal tissue, cells are only able to divide and move about in very restricted and organized ways. But when there is a wound, basically, cells set aside what they normally do and their main goal becomes closing up this defect in the tissue," explains Dr. Chang, assistant professor of dermatology at Stanford University School of Medicine, Stanford, Calif.

"The cells that are normally not dividing are now growing fast. They cross various tissue planes to migrate towards the wound, bring in new blood vessels and close up the defect."

Wound healing/cancer similarities

Some of these same steps, including cell division and ability to migrate across tissue planes, are useful properties for cancer cells.

A strong parallel between wound healing and cancer formation in dermatology is seen with epidermolysis bullosa in pediatric patients. These children's genetic defect not only causes numerous chronic wounds, but also puts them at high risk for developing skin cancer at young ages, Dr. Chang tells Dermatology Times.

Based on this and other evidence, Dr. Chang and colleagues set out to determine whether they could identify a set of genes involved in wound healing. They began looking for that commonality in cancer samples and using it to predict the course of cancer.

Finding the genes

The researchers used gene array technology. The measurement of gene activity was observed simultaneously for all genes in the genome, creating a "snap shot."

"We took fibroblasts and exposed them to serum, which is a fraction of clotted blood, as a way of modeling an encounter that would happen during wound healing," Dr. Chang says. "We then just monitored the changes in the global gene expression pattern of these cells."

They discovered a pattern of 512 genes that were indicative of the serum response in their simple wound-healing model. These genes would reproducibly change in the wound-healing environment. The researchers found that, in many kinds of cancers, that same pattern of 512 genes was turned on.

"We actually showed that in several kinds of cancers, especially breast cancer, this is a very useful predictor of outcomes in early cancer patients. Cancers that have these wound healing genes tend to do much worse," he said.

Dr. Chang et al. analyzed a database of tumor gene activities profiles of 295 breast cancer patients in the Netherlands and classified study participants as having either activated or quiescent wound-response signatures. Of the group, 126 had activated wound-response signatures and 169 tumors were classified as quiescent.

According to the study, participants in the activated group were three times more likely to die within 10 years of diagnosis. While half of the participants in the activated group had died within the decade, only 16 percent of participants in the quiescent group had died in the same time period. Activated tumors were twice as likely as quiescent tumors to recur with a distant metastasis.

A new predictive tool

The novel finding was independent of the known predictors, or clinical risk factors, in cancer today.

"If this were added to the current knowledge, it would actually improve the stratification of patients based on their future risk," according to Dr. Chang.

The researchers say that the gene signature response would be most useful, clinically, in deciding if an early stage cancer patient should get chemotherapy.