SAN ANTONIO – Diagnosis of diabetes within the prior 4 years was independently associated with breast cancer in a Swedish case-control study.
Dr. Håkan Olsson reported at the San Antonio Breast Cancer Symposium on all 2,724 women diagnosed with breast cancer in southern Sweden during 2005-2007 and 20,542 matched controls. He and his coworkers were interested in how the malignancy is related to diabetes, obesity, and serum lipid levels.
Dr. Hakan Olsson
In a multivariate analysis adjusted for obesity, serum lipids, and other potential confounders, the breast cancer patients were 37% more likely than controls to have been diagnosed with diabetes during the previous 4 years. Yet diabetes diagnosed 4-10 years previously was not associated with a significant increase in breast cancer, said Dr. Olsson, professor of oncology and cancer epidemiology at Lund (Sweden) University.
The most likely explanation for the finding that only relatively recently diagnosed diabetes was linked to increased risk of breast cancer is that the diabetic hormonal milieu doesn’t initiate breast tumors, but rather it promotes the growth of tumors that are already established but dormant. This would be analogous to the relationship between hormone replacement therapy and breast cancer, where the Early Breast Cancer Trialists’ Collaborative Group has demonstrated that it’s only present use, not past use, that increases the risk of malignancy, he observed.
Dr. Olsson and coworkers also looked at the relationship between diabetes and all other types of cancer among patients in the comprehensive regional cancer registry. They found that three other types of cancer in addition to breast cancer were associated with a significantly increased likelihood of prior diagnosis of diabetes compared to controls: pancreatic cancer, with a 2.36-fold increased rate; liver cancer, with a 3.43-fold increased odds; and colon cancer, with a 1.49-fold increase.
Dr. Olsson and coworkers also looked at the relationship between the antidiabetic medications metformin and glargine and cancer risk among all patients in the cancer registry, not just those with breast cancer. Use of the long-acting insulin analog glargine, which is quite common among Swedish type 2 diabetic patients, was associated with a 2.88-fold increased overall cancer risk. In contrast, metformin use was associated with an 8% reduction in overall cancer risk, although this association didn’t achieve statistical significance, unlike the relationship between glargine and overall cancer.
An association between glargine and increased cancer risk has also been noted in several other studies, according to Dr. Olsson.
In the southern Swedish breast cancer cohort, obesity after age 60 was independently associated with a 55% increased likelihood of breast cancer after controlling for diabetes and other factors in a multivariate analysis. However, obesity in women under age 60 was associated with a nonsignificant 41% reduction in breast cancer.
To Dr. Olsson’s surprise, the investigators found that hypercholesterolemia was independently associated with a 27% reduction in the prevalence of breast cancer. In other words, significantly fewer breast cancer patients had a high cholesterol level compared with the general population. This is a novel finding that requires confirmation in other data sets, he added.
The relationship between metformin and breast cancer is under study in the large, prospective, phase III, double-blind, randomized MA 32 clinical trial led by the National Cancer Institute of Canada Clinical Trials Group. More than 1,000 nondiabetic patients with early-stage breast cancer have been randomized to metformin or placebo. Key end points in the MA 32 study include overall and disease-free survival. Results are about 5 years off.
Dr. Olsson’s study was funded by the Southern Sweden Health Care Region. He declared having no relevant financial relationships.