When Cancer Runs in the Family

Most cancers occur sporadically. At least 5.5 percent of cancer cases are based on a genetic predisposition. These are usually identified because first-degree relatives develop the same type of tumor. So far, scientists have presumed that only a few types of tumor have such a familial form. However, Professor Kari Hemminki of the Division of Molecular-Genetic Epidemiology of the Deutsches Krebsforschungszentrum (German Cancer Research Center, DKFZ), has now found familial forms of almost all tumor types. In the families affected, both the offspring and siblings of a cancer sufferer are at an increased risk to develop the same disorder.

Sweden with its extensive population statistics offers unique opportunities to study the genetic risks of cancer. The Swedish “family register", a record of all individuals born in Sweden after 1932 and their parents, contains more than 10 million individuals. Moreover, the Swedish Cancer Register records almost 100 percent of cancer cases occurring in the population. Integrating both databases, which are also used at DKFZ, Hemminki and his co-workers were able to analyze over three million families. He identified almost 5000 families in which several cases of the same type of tumor occurred – an indicator of familial cancer. Based on these data, epidemiologist Hemminki was able to calculate the familial risks for each tumor type with unprecedented accuracy.

Hemminki found familial forms in 24 out of 25 tumor types. Familial tumors were most frequent in prostrate cancer at about 15% of cases, followed by intestinal cancer (10%) and breast cancer (8.5%). Familial cancers were least frequent in connective tissue tumors (0.4%) and testicular tumors (0.5%).
The degree of familial clustering of cancer also varies considerably among the different types of cancer. The highest genetic risk was found in families with testicular cancer. The sons of affected fathers had a fourfold increased risk compared to sons of families without testicular cancer, the brothers of affected individuals even had a ninefold increased risk to develop this cancer.

As opposed to testicular cancer, familial forms of Hodgkin’s disease, a type of lymphoma, had been unknown to date. Hemminki identified familial cases of this type of tumor, which also exhibited a strong familial risk: In children of affected persons the risk was increased almost fivefold, in siblings even sixfold. A strong genetic component was also revealed in families with non-medullary thyroid cancer, esophageal cancer and a certain type of bone cancer (multiple myeloma).

The more common cancer types such as cancers of the prostate, kidney, skin, stomach and lung, leukemias and endocrinal tumors also occur in a familial form, even more frequently than previously thought. Hemminki observed particulary high risks to develop the disease in families with more than two cases in first-degree relatives and/or early age of onset. Both indicators suggest a strong genetic predisposition to the respective cancer and should possibly give cause for genetic counseling. In general, Hemminki’s findings show that current genetic counseling practice covers only a fraction of familial cancers.

Hemminki rejects the argument that the observed familial clustering of cancer cases could also be caused by shared environmental factors. In previous studies with married couples he had shown that a parallel increase in cancer risks only occurred in tumors that are strictly environmentally caused (e.g., lung or genital cancers). However, the risk usually remains far below that of genetically caused tumors.

*Kari Hemminki, Xinjun Li, Kamila Czene: Familial risk of cancer: Data for clinical counselling and cancer genetics; International Journal of Cancer, Volume 108, Issue 1, 109-114