番茄社区

SAN ANTONIO -- Clinical trials of the diabetes drug metformin in breast cancer yielded a few possible clues to anticancer activity but little else, according to a series of reports.

The largest of three trials showed no significant effect on tumor-cell apoptosis in women who received metformin prior to surgery. Higher apoptotic activity was observed in patients without insulin resistance compared with the insulin-resistant subgroup.

Results of the other two trials hinted at host-specific factors that might influence the drug's effect on apoptosis and cell proliferation, as reported at the San Antonio Breast Cancer Symposium.

In a review of the three studies and others included in a poster session, Michael N. Pollak, MD, borrowed and characterized the status of metformin research in cancer as "approaching the end of the beginning."

"We're not really yet at the middle or concluding phases of the research, because there are a lot of unanswered questions, and some of those questions have become clear from the review of the posters today," said Pollak, of McGill University in Montreal.

The rationale for investigating the anticancer potential of metformin has as its origin observations that diabetic patients treated with metformin had a lower prevalence of cancer compared with diabetic patients treated with other medications or no medication.

Laboratory studies have suggested multiple mechanistic pathways by which metformin might inhibit tumor growth or proliferation, including Stat3, AMP kinase, and insulin receptor kinase, said Pollak. However, studies involving laboratory models also have resulted in higher metformin concentrations than those achieved with conventional dosing in patients.

"Insulin -- or something correlated with insulin -- influences prognosis, but early trials have shown perturbations of insulin by metformin that are small in magnitude and that vary between subjects," said Pollak. "Other less-explored candidate metformin effects on the host include changes in inflammatory cytokines, serum kinase-activating activity, and others."

Moreover, placebo-controlled studies have demonstrated unexpected increases in the proliferation marker Ki-67 between baseline and surgical biopsies in placebo-treated patients. The findings might represent an artifact or suggest that surgery somehow influences tumor proliferation, Pollak continued.

As described in one of the poster presentations, Italian investigators evaluated cancer-cell apoptosis in 88 breast cancer patients who were randomized to metformin or placebo for 4 weeks prior to surgery for early breast cancer. Apoptosis was assessed by TUNEL assay in biopsy samples obtained before randomization and immediately prior to surgery.

Apoptosis was similar between groups at baseline (4% and 3% in the metformin and placebo arms, respectively) and increased significantly (P<0.0001) in both arms prior to surgery (10% and 8%, respectively). TUNEL levels at surgery did not differ significantly between groups, according to Giancarlo Pruneri, MD, of the European Institute of Oncology in Milan.

Investigators found borderline-significant evidence (P=0.1) of interaction between TUNEL and homeostasis model assessment (HOMA), a measure of insulin resistance. Patients who were not insulin resistant (HOMA <2.8) had a TUNEL level of 10% with metformin versus 6% with placebo (P=0.05).

In contrast, women with insulin resistance (HOMA ≥2.8) had a median TUNEL value of 6% with metformin and 9% with placebo (P=0.3).

The data also showed significant correlation between the proliferation factor Ki-67 and TUNEL at baseline and surgery (P<0.0001).

Another small presurgical study showed no effect of metformin on tumor proliferation, as reflected by levels of Ki-67. The study involved 35 overweight and obese women (body mass index [BMI] >25) with early breast cancer (N=25) or ductal carcinoma in situ (N=10).

The patients started metformin 2 to 4 weeks before surgery (median duration 22 days), and the study's primary endpoint was the change in Ki-67 activity from baseline to surgery, Kevin Kalinsky, MD, of Columbia University in New York City, and colleagues reported.

The mean Ki-67 value did not change significantly in patients treated with metformin or among 58 similar patients who did not receive metformin prior to surgery. Treatment with metformin did lead to significant reductions in cholesterol (P<0.01) and leptin (P=0.03) and a trend toward reductions in insulin levels, HOMA, and adiponectin.

Results of a small Canadian study showed a consistent, potentially beneficial effect of metformin in women with early breast cancer. The study involved 39 nondiabetic women with newly diagnosed breast cancer. They received metformin for at least 2 weeks prior to diagnostic core biopsy.

Neoadjuvant metformin was associated with a significant (P<0.05) reduction in BMI, weight, glucose, and HOMA, and trend toward lower insulin levels. Ki-67 activity decreased significantly (P=0.0160), and TUNEL increased significantly (P=0.0037).

Insulin receptor expression decreased significantly (P=0.0375), as did AMPK (P=0.0034) and Akt signaling (P=0.0001), which are associated with tumor proliferation.

Taken together, the findings were consistent with beneficial anticancer effects of metformin, Ryan J.O. Dowling, PhD, of the Ontario Cancer Institute in Toronto, and colleagues reported. In particular, changes in insulin levels (as reflected in increased insulin receptor expression) and Akt signaling suggest metformin has clinically important insulin-dependent effects on tumor growth.

A second report based on the Canadian study delved into the details of biomarker analyses performed on tumor specimens. The primary objective was to gain clarifying mechanistic information to support metformin's anticancer activity.

The results were "compatible with the concept that metformin works in vivo via upregulation of tumor phosphorylated AMPK and down-regulation of phosporphylated Akt and proliferation," Sirwan M. Hadad, MD, of the University of Sheffield in England. "Since down-regulation of pAMPK is a feature of breast cancer, this suggests mechanistic evidence for the therapeutic effect of metformin."