Use of diabetes drugs, especially metformin, prior to surgery for prostate cancer is associated with an increased risk of high-grade tumors, investigators reported.
In a study of 1314 PCa patients who underwent radical prostatectomy, a team led by Teemu Murtola, MD, PhD, of Tampere University Hospital in Tampere, Finland, found that preoperative use of diabetes drugs was associated with 83% increased odds of Gleason 7–10 disease compared with non-use, according to a paper published online ahead of print in Prostate Cancer and Prostatic Diseases. Preoperative use of metformin was associated with 3-fold greater odds of high-grade PCa compared with preoperative use of other diabetes drugs. Despite this, DrMurtola’s group found no difference in PSA levels or pathologic stage between metformin users and uses of other diabetes drugs. Users of diabetes drugs had a decreased risk of pathologic stage T3-T4 and/or N1 in an analysis adjusted for HbA1c. The study found no association between insulin use and Gleason grade, pathologic stage, or PSA level.
During a median follow-up of 8.6 years after surgery, 551 men experienced biochemical recurrence and 244 died, 32 from PCa. On multivariable analysis, preoperative use of diabetes drugs was associated with an 81% increased risk of death due to any cause compared with non-use. The investigators reported that, generally, they observed no association between risk of disease recurrence and preoperative use of diabetes drugs. In stratified analysis, preoperative use of diabetes drugs was associated with an increased risk of disease recurrence in men with low-grade disease, but not among those with Gleason 7–10 disease.
Men who discontinued diabetes drugs after surgery had a nearly 4.7-fold increased risk of death compared with non-users, whereas no risk difference was observed in those who kept using the drugs. “This probably reflects the patients’ increased likelihood to stop preventive medication use at terminal phase of cancer,” the investigators explained.
Despite the increased frequency of high-grade tumors among users of diabetes drugs, the investigators found no difference in median PSA level at diagnosis. “This supports the idea that reliability of PSA as a tumor marker might be lower among diabetics compared with non-diabetics,” they wrote. “However, this did not have impact on risk of PSA recurrence.”
Dr Murtola and his colleagues noted that important strengths of the study included the availability of comprehensive information on disease characteristics and progression, detailed information on medication use, and information on glycemic control among users of diabetes drugs. “We were able to analyze the effect of drug use both before and after surgery.”
The investigators acknowledged study limitations, including a low number of PCa deaths, an inability to differentiate between types of diabetes mellitus, and lack of information on diet management.